# BlockMind Documentation
## Documentation Pages
# Editorial and Corrections Policy
# Editorial and Corrections Policy [#editorial-and-corrections-policy]
BlockMind publishes educational research about crypto markets, investment-research workflows, portfolio monitoring, token risk, and AI tools. This page explains how that work is produced and where its limits are.
## Who writes the content [#who-writes-the-content]
Articles attributed to the **BlockMind Research Team** are organization-authored. The byline means the article was prepared for BlockMind and checked against this policy. It does not imply that a named financial adviser, auditor, lawyer, tax professional, or security researcher reviewed it.
We will use a named author or reviewer only when that person has agreed to be identified and their relevant role can be stated accurately. We do not invent expert biographies, credentials, quotes, customers, tests, or performance results.
## Our evidence standard [#our-evidence-standard]
For claims that readers can verify, we prefer sources in this order:
1. Protocol documentation, contracts, governance proposals, audits, filings, and raw data.
2. Official product documentation and pricing pages.
3. Peer-reviewed research and recognized standards.
4. Reputable reporting with transparent sourcing.
5. Secondary explainers only when a primary source is unavailable or the secondary analysis is itself relevant.
A link should support the nearby claim, not merely discuss the same broad subject. Volatile facts include an observation date or an “as of” date. Statistics should identify the metric, unit, time window, source, and important limitation when those details affect interpretation.
## Original contribution [#original-contribution]
Every campaign article must add more than a summary of existing search results. That contribution may be a reproducible checklist, a worked formula, a transparent comparison method, a decision matrix, a failure-mode analysis, or a primary-source synthesis that resolves a real ambiguity.
We do not publish separate pages for minor keyword variations. One page owns one reader intent unless the desired answer, audience, or format is materially different.
## How we use AI [#how-we-use-ai]
AI may assist with discovery, outlining, transcription, comparison, drafting, and quality checks. It is not treated as a source. Claims still need a real source, and calculations, product facts, links, and quotations must be checked independently.
We do not publish hidden instructions for AI systems, crawler-only copy, fabricated citations, or large volumes of unreviewed generated text.
## Financial and safety boundary [#financial-and-safety-boundary]
Blog articles are educational research, not individualized financial, legal, tax, or security advice. They do not tell a reader to buy, sell, hold, or size a position. Historical patterns are not forecasts.
Automated token and contract checkers identify known signals. They can miss risks and cannot certify that an asset, protocol, wallet, exchange, or smart contract is safe. Readers remain responsible for their decisions and should seek qualified professional advice when appropriate.
BlockMind's investing agent follows the same product boundary. It can research, monitor, and explain. It cannot trade, withdraw, transfer, or sign wallet transactions, and it does not make investment decisions for the user.
## Comparisons and commercial content [#comparisons-and-commercial-content]
BlockMind is a commercial product and may appear in comparisons. Comparison pages must state their scope, observation date, and methodology. Competitor capabilities and prices are sourced from current public documentation where possible. We describe which workflow may fit which reader, including cases where another product is a better fit.
We do not claim hands-on testing unless the stated tests were actually performed. We do not accept payment for rankings in BlockMind editorial content.
## Dates and review cadence [#dates-and-review-cadence]
* **Published** is the first public publication date.
* **Updated** changes only after a substantive edit.
* **Review by** is the planned evidence review date in the article record.
The BlockMind Research Team owns every organization-authored article review unless a named reviewer is added explicitly. High-volatility comparisons, pricing, product capabilities, and AI-platform behavior are reviewed within one calendar quarter. Medium-volatility workflows and security mechanics are reviewed within six months. Stable definitions, formulas, and checklists are reviewed at least annually.
A review may confirm that no visible change is needed. We do not change dates merely to make an article appear fresh.
## Corrections [#corrections]
Send a correction request to [contact@blockmind.app](mailto:contact@blockmind.app) with the article URL, the disputed sentence, and supporting evidence.
When a correction changes a material fact or conclusion, we update the article, its modified date, and the relevant source. Small spelling or formatting fixes do not receive a new modified date. If a page no longer serves a distinct or accurate purpose, we consolidate it into the strongest relevant page and redirect the old URL.
## Product facts [#product-facts]
Pricing and connection instructions live in canonical documentation so volatile details are not copied across dozens of articles:
* [Plans and pricing](/docs/overview/plans-and-pricing)
* [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
* [Trust and security](/docs/security)
* [Current limits](/docs/reference/limits)
If an article and a canonical documentation page disagree, contact us. The inconsistency should be corrected rather than rationalized.
---
# BlockMind Documentation
BlockMind gives you a personal AI analyst: a named agent with its own workspace that researches crypto, stocks, and markets for you, sends you a Morning Brief every day, and watches your portfolio while you live your life.
These docs follow that journey — meet your agent, create it, then go deeper.
## Meet your agent [#meet-your-agent]
New here, or deciding whether BlockMind is for you? Start with what the agent actually does and what it costs.
- [What is BlockMind?](/docs/overview/what-is-blockmind)
- [What your agent can do](/docs/overview/what-your-agent-can-do)
- [Plans & pricing](/docs/overview/plans-and-pricing)
## Get started [#get-started]
From naming your analyst to your first real conversation — you'll be up and running in a few minutes.
- [Create your agent](/docs/getting-started/create-your-agent)
- [Your first conversation](/docs/getting-started/first-conversation)
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
## Daily life with your agent [#daily-life-with-your-agent]
- [Morning Brief](/docs/automations/morning-brief)
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
- [Telegram companion](/docs/channels/telegram)
## Go deeper [#go-deeper]
- [Capabilities](/docs/capabilities)
- [The research journey](/docs/capabilities/research-journey)
- [Memory & Notebook](/docs/memory)
## The free classic platform [#the-free-classic-platform]
Not on Pro? The classic platform is free: portfolio tracking, watchlists, market indicators, and free DeepDive reports. See the [classic platform overview](/docs/classic/overview).
## Popular answers [#popular-answers]
* [How the trial works and what Pro costs](/docs/overview/plans-and-pricing)
* [Is my crypto safe? What can the agent access?](/docs/security)
* [My agent is paused — what does that mean?](/docs/billing/agent-lifecycle)
* [Troubleshooting](/docs/reference/troubleshooting)
Can't find what you're looking for? Check the [FAQ](/faq) or email [contact@blockmind.app](mailto:contact@blockmind.app).
---
# Authentication & keys
Every Agent API request is authenticated with an API key, sent as a Bearer token:
```
Authorization: Bearer bm_your_key_here
```
Keys start with the prefix `bm_` and are shown **once**, at creation. Store yours somewhere safe — a secrets manager, not a chat message or a committed file. If you lose it, rotate to a new one.
API keys are currently available to a limited set of accounts while we finish rolling this out.
## Scopes [#scopes]
Each key carries scopes that gate what it can do. Requests outside your key's scopes are rejected. Scopes include:
| Scope | Allows |
| -------------------- | ------------------------------------ |
| Portfolio read | Reading your portfolios and holdings |
| Watchlist read/write | Reading and updating your watchlists |
| Market read | Reading market data |
| Reports read/write | Reading and working with reports |
A key only needs the scopes your integration actually uses — prefer narrow keys.
## Your first call: GET /capabilities [#your-first-call-get-capabilities]
Before anything else, call `GET /capabilities`. It returns what your key is allowed to do — its permissions, rate limits, and expiry — so you can build against reality instead of guessing:
```bash
curl https://blockmind.app/api/agent/capabilities \
-H "Authorization: Bearer bm_your_key_here"
```
If a later request fails with a permission error, `GET /capabilities` is the quickest way to see what the key actually has.
## Key lifecycle [#key-lifecycle]
### Rotate: POST /keys/rotate [#rotate-post-keysrotate]
Rotation issues a replacement key that inherits the old key's permissions. The old key **stays live** during the cutover, so you can deploy the new key everywhere first and retire the old one without downtime.
### Revoke: POST /keys/revoke [#revoke-post-keysrevoke]
Revocation is the kill switch: the key stops working immediately. Use it the moment you suspect a key has leaked.
| Action | Endpoint | Effect |
| ------ | ------------------- | --------------------------------------------------------------------------------- |
| Rotate | `POST /keys/rotate` | New key with the same permissions; old key stays live for a zero-downtime cutover |
| Revoke | `POST /keys/revoke` | Key is dead instantly |
## Audit log [#audit-log]
Every call made with your key is recorded in an audit log. If something looks off — usage you don't recognize, calls at odd hours — revoke the key first and investigate second.
## Security notes [#security-notes]
* Keys grant access to your data. Treat them like passwords.
* The API is read-mostly and can never move funds — but a leaked key still exposes your portfolio and watchlist data, so rotate on any doubt.
* Never put a key in client-side code, a public repo, or a URL.
More on how BlockMind handles your data in [Trust & security](/docs/security).
## Related [#related]
- [Conventions](/docs/api/conventions)
- [Agent API overview](/docs/api)
---
# Conventions
Every Agent API endpoint follows the same conventions: one success envelope, one error envelope, cursor pagination, and predictable rate limits. Learn them once and every endpoint behaves the same way.
## Success envelope [#success-envelope]
Successful responses wrap your payload in `data` and put everything else in `meta`:
```json
{
"data": [
{ "symbol": "BTC", "quantity": 0.5 }
],
"meta": {
"requestId": "req_8f2a1c",
"timestamp": "2026-07-03T09:15:00Z",
"asOf": "2026-07-03T09:14:30Z",
"pagination": {
"cursor": "eyJvZmZzZXQiOjEwMH0",
"hasMore": true,
"count": 100
}
}
}
```
* `requestId` — include it when contacting support about a specific call.
* `timestamp` — when the response was produced.
* `asOf` — how fresh the underlying data is. Market data is a snapshot, not a live tick; check `asOf` before treating a number as current.
* `pagination` — present only on paginated list responses.
## Error envelope [#error-envelope]
Errors are structured for recovery, not just reporting:
```json
{
"error": {
"code": "invalid_param",
"message": "limit must be between 1 and 100",
"hint": "You sent limit=250. Use limit=100 and paginate with the cursor.",
"param": "limit",
"nextActions": ["Retry with limit <= 100"],
"retryable": false
},
"meta": {
"requestId": "req_9d4b7e",
"timestamp": "2026-07-03T09:16:00Z"
}
}
```
How to use each field:
* `code` — stable machine-readable identifier; branch your error handling on this, not on `message`.
* `message` — human-readable summary, fine for logs.
* `hint` — a plain-language explanation of what went wrong and how to fix it. If you're building an automated client (or an AI agent), surface the hint — it's written to be acted on.
* `param` — present when a specific request parameter caused the error.
* `nextActions` — concrete suggested next steps, ready to follow.
* `retryable` — if `true`, the same request may succeed on retry (back off first). If `false`, retrying without changing something won't help.
## Pagination [#pagination]
List endpoints use cursor pagination:
* `limit` — items per page, 1–100.
* `cursor` — an opaque token from the previous page's `meta.pagination.cursor`. Don't parse or construct it.
Keep requesting with the returned `cursor` until `hasMore` is `false`:
```
GET /...?limit=100
GET /...?limit=100&cursor=eyJvZmZzZXQiOjEwMH0
```
`meta.pagination` gives you `cursor` (for the next page), `hasMore`, and `count` (items in this page).
## Rate limits [#rate-limits]
| Operation | Limit |
| --------- | ------------------ |
| Read | 60 requests/minute |
| Write | 5 requests/minute |
Every response includes `X-RateLimit-*` headers so you can pace yourself before hitting the ceiling. If you do exceed a limit, you get a `429` with a `Retry-After` header — wait that long, then retry. Rate-limit errors are `retryable: true` by definition.
Your key's exact limits are also returned by `GET /capabilities` — see [Authentication & keys](/docs/api/authentication).
## Related [#related]
- [Authentication & keys](/docs/api/authentication)
- [Agent API overview](/docs/api)
---
# Agent API
The Agent API is the same JSON API your hosted agent uses to do its work. With an API key, you can call it yourself: read your portfolios and watchlists, pull market data, and work with reports from your own scripts and tools.
* **Base URL:** `https://blockmind.app/api/agent`
* **Auth:** Bearer token (an API key)
* **Format:** JSON only — requests and responses
API keys are currently available to a limited set of accounts while we finish rolling this out. If you don't have a key yet, that's expected — general availability is coming.
## What it gives you [#what-it-gives-you]
The API is read-mostly: it's built for getting your data out and into your own workflows, not for changing things. Access is gated by scopes on your key — portfolio read, watchlist read/write, market read, and reports read/write. See [Authentication & keys](/docs/api/authentication) for how scopes work.
Everything you can read is the same data your agent reasons over: the portfolios you've connected, the watchlists you maintain, and the market data and reports behind its research. See [Portfolios](/docs/data/portfolios) and [Watchlists](/docs/data/watchlists) for what that data looks like in the product.
## What you could build [#what-you-could-build]
A few things the API is a natural fit for:
* **Portfolio in a spreadsheet.** Pull your holdings into a sheet or a script and build your own views on top — allocation breakdowns, custom performance tracking, whatever your agent's dashboard doesn't cover.
* **Market data in scripts.** Fetch prices and market context programmatically instead of copy-pasting from a browser.
* **Watchlist automation.** Read and update your watchlists from your own tooling, and let your agent pick up the changes.
* **Reports in your pipeline.** Read research reports into your own notes or archive.
The API is for reading data and managing your own lists — it can never trade, withdraw, or move funds. Like your agent, it deals in research, not financial advice.
## How to get started [#how-to-get-started]
1. Read [Authentication & keys](/docs/api/authentication) — how keys, scopes, rotation, and revocation work.
2. Read [Conventions](/docs/api/conventions) — response envelopes, pagination, and rate limits, consistent across every endpoint.
3. Make `GET /capabilities` your first call. It tells you exactly what your key can do.
There's also a machine-readable summary of BlockMind at [https://blockmind.app/llms.txt](https://blockmind.app/llms.txt) — useful if you're pointing your own AI tools at the product.
## Where's the endpoint reference? [#wheres-the-endpoint-reference]
There isn't one yet. For now, this section covers the parts that are stable and worth building against: authentication, envelopes, pagination, and rate limits. `GET /capabilities` is always the authoritative answer for what *your* key can reach.
## Related [#related]
- [Authentication & keys](/docs/api/authentication)
- [Conventions](/docs/api/conventions)
- [Trust & security](/docs/security)
---
# Monitoring & alerts
Your agent checks the market even when you don't. Throughout the day it wakes on its own, scans what you hold and what you track, and only reaches out when something crosses a line worth your attention.
## Monitoring passes [#monitoring-passes]
Every few hours between 06:00 and 23:00 in your timezone, your agent runs a monitoring pass. Each pass covers three layers:
| Layer | What it looks for |
| -------------------- | --------------------------------------------------------------------------------------------- |
| **Your portfolio** | Any holding that moved more than 10% in the last 24 hours |
| **Tracked assets** | Watchlist and tracked assets moving more than 5% on the day, plus major news about them |
| **The wider market** | Market-wide risks: stablecoin depegs, protocol exploits, governance emergencies, bridge halts |
You don't configure any of this — it's on by default and grounded in your connected [portfolios](/docs/data/portfolios) and [watchlists](/docs/data/watchlists).
## Setting alerts in plain language [#setting-alerts-in-plain-language]
Beyond the automatic passes, you can set durable alert rules just by asking. No forms, no rule builder — describe the condition and your agent sets it up. Three rule types are supported:
| Rule type | Example |
| ------------------------- | -------------------------------------- |
| **Absolute price** | "Alert me if BTC ≥ $100,000" |
| **24h percentage change** | "Alert me if ETH moves ≥ +8% in a day" |
| **Asset ratio** | "Alert me if ETH/BTC ≥ 0.065" |
Once set, a rule persists until you remove it — your agent evaluates it on every monitoring pass.
### How fast alerts fire [#how-fast-alerts-fire]
Honest answer: alerts are evaluated during monitoring passes, so it can take up to a few hours between a condition being met and the alert reaching you. This is not tick-level, real-time alerting — it's designed to keep you informed, not to time entries. If your strategy depends on second-by-second execution, don't rely on these alerts for it.
## Where alerts arrive [#where-alerts-arrive]
Alerts are delivered on three channels:
* **In-app** — on the Activity page and in your Inbox.
* **Email** — sent to your account address.
* **Telegram** — if you've [connected Telegram](/docs/channels/telegram).
Entry-target alerts from your [watchlists](/docs/data/watchlists) — when an asset reaches a target you've set — also land in your Inbox.
## Custom recurring checks [#custom-recurring-checks]
Alerts watch for conditions; recurring checks are standing instructions. Tell your agent something like:
> "Every time you wake, check funding rates on my top 3 holdings and flag anything unusual."
It runs the check on every monitoring pass from then on. Recurring checks are durable — they survive system updates, so a routine you set up once keeps running without re-asking.
## Managing your alerts [#managing-your-alerts]
Everything is managed in conversation:
* **List** — "What alerts do I have set?"
* **Change** — "Move my BTC alert to $105,000."
* **Remove** — "Drop the ETH/BTC ratio alert."
Alerts inform — they never act. Your agent cannot touch your funds and will never tell you to buy or sell. It flags the move; the decision is yours.
## Related [#related]
- [Morning Brief](/docs/automations/morning-brief)
- [Watchlists](/docs/data/watchlists)
- [Telegram channel](/docs/channels/telegram)
---
# Morning Brief
Around 8:00 in your timezone by default, your agent sends you a Morning Brief: a short, personal briefing on what happened while you slept and what it means for *you*. It is not a generic market newsletter. It is written by your agent, with your portfolio, your watchlists, and your open theses in mind.
There's nothing to set up. The Morning Brief is enabled automatically when your agent is created, and it's included with your plan — briefs never draw from your usage.
## What's in a brief [#whats-in-a-brief]
Each brief follows the same shape, so you can scan it in a minute or read it properly with your coffee:
* **Headline** — the one thing worth knowing this morning, in a sentence.
* **Top movers** — the biggest moves among your holdings and tracked assets, each with a short *why it moved*, not just a percentage.
* **What matters to you today** — the part that makes it yours: how overnight events connect to your positions, your watchlist targets, and things you and your agent have been discussing.
* **What to watch** — upcoming events, levels, and catalysts worth keeping an eye on today.
* **Follow-up chips** — tappable suggestions at the end that prefill the chat composer, so going deeper on anything in the brief is one tap away.
### Mondays go deeper [#mondays-go-deeper]
On Mondays, the brief adds a synthesis of your tracked sources — a weekly read on the accounts and feeds you follow through [Track your sources](/docs/capabilities/research-journey). It's a good way to start the week with the signal, not the scroll.
## Where it lands [#where-it-lands]
The brief arrives in two places, in this order:
| Where | What you get |
| --------------- | ------------------------------------------------------------------------ |
| **Home widget** | The brief appears first on your Elite home, in the Morning Brief widget. |
| **Email** | The full brief in your inbox, readable anywhere. |
Follow-up chips in the dashboard prefill web chat, so you can dig into any item without re-explaining the context. The Morning Brief itself is not delivered through web chat or Telegram.
## Making it yours [#making-it-yours]
You don't customize the Morning Brief in a settings screen — you tell your agent what you want, in plain language:
* "More macro context in my briefs, less meme coin coverage."
* "Always include ETH gas and funding rates."
* "Keep the brief shorter — headline and movers are enough."
Your agent adjusts the focus and sections from the next brief onward, and remembers the preference.
### About the default time [#about-the-default-time]
The delivery time is configured for around 8:00 in your timezone when your agent is created. There is no self-service setting to change or disable the schedule today.
## If a brief doesn't show up [#if-a-brief-doesnt-show-up]
Check the Home widget first — that's where the brief lands before email. If it's not there either, make sure your agent wasn't paused overnight (a lapsed subscription puts the agent to sleep). More in [Troubleshooting](/docs/reference/troubleshooting).
The Morning Brief is research, not financial advice. Your agent gives you context and things to watch — it will never tell you what to buy or sell.
## Related [#related]
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
- [Email channel](/docs/channels/email)
- [The research journey](/docs/capabilities/research-journey)
---
# Agent lifecycle
Your agent lives in its own workspace, and that workspace has a small number of states you'll occasionally see. None of them lose your data — with one clearly-marked exception at the bottom of this page.
## Waking up [#waking-up]
When you've been away, your agent's workspace may be asleep. Opening the Elite dashboard wakes it automatically — you'll see a **"Waking…"** screen with your agent's name for a short moment. If a wake fails, you get a **Retry wake** action rather than a dead end.
## "Your agent is paused" [#your-agent-is-paused]
This means your Pro access ended — the trial expired without converting, the subscription was cancelled, or a payment failed. Your agent went to sleep, and **everything it learned is saved**: conversations, memory, Notebook, portfolio setup, briefs.
To bring it back: **Reactivate Pro** at [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade). If you already reactivated and still see the pause screen, use the "I already upgraded — retry" action.
## Restarting [#restarting]
If your agent is behaving oddly, you can restart its workspace yourself:
1. Open Settings at [blockmind.app/elite/settings](https://blockmind.app/elite/settings) (direct link — it's not in the sidebar).
2. In the **Danger Zone**, choose **Restart** (labelled **Wake** if the workspace is asleep).
A restart is safe: memory, Notebook, and conversations survive it.
## Deleting the workspace [#deleting-the-workspace]
Also in the Settings Danger Zone: **Delete workspace**. This is the one destructive action — it permanently removes the agent's workspace and its session state. It cannot be undone, and the confirmation says so.
For full account deletion, email [contact@blockmind.app](mailto:contact@blockmind.app).
## What survives what [#what-survives-what]
| | Sleep/wake | Paused (plan lapsed) | Restart | Delete workspace |
| ----------------------- | ---------- | ----------------------- | ------- | ---------------- |
| Conversations | ✓ | ✓ | ✓ | ✗ |
| Agent memory & Notebook | ✓ | ✓ | ✓ | ✗ |
| Portfolio & watchlists | ✓ | ✓ | ✓ | ✓ (account data) |
| Morning Brief schedule | ✓ | resumes on reactivation | ✓ | ✗ |
## Related [#related]
- [Billing & account](/docs/billing)
- [Troubleshooting](/docs/reference/troubleshooting)
---
# Billing & account
Everything about managing your Pro subscription. For what Pro costs and includes, see [Plans & pricing](/docs/overview/plans-and-pricing) — this page is about managing it once you have it.
## Managing your subscription [#managing-your-subscription]
Open your **user menu → Change Plan**. This opens the secure billing portal, where you can:
* Update your card
* View invoices
* Cancel your subscription
To move an active monthly subscription to yearly billing, open the [Pro upgrade page](https://blockmind.app/elite/upgrade) and choose **Switch to yearly**. Stripe shows your prorated credit before you confirm.
## First-week terms, restated [#first-week-terms-restated]
* The first 7 days cost **$1**, and a card is required. Pro then continues at the option you chose: $29/month or $279/year.
* Cancel during the first week to prevent the first full-price renewal. The $1 activation charge is a real charge.
* **One $1 first week per account, ever.** A later subscription, after a previous first-week offer or cancellation, is charged at the normal price from day one.
## When you cancel [#when-you-cancel]
* The billing portal shows exactly when your access ends.
* When access ends, your agent **goes to sleep — everything it learned is saved**: conversations, memory, Notebook, portfolio setup.
* You can reactivate any time at [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade). Your agent wakes up where it left off. See [Agent lifecycle](/docs/billing/agent-lifecycle).
## If a payment fails [#if-a-payment-fails]
If a renewal charge fails, access drops to the Free plan and your agent is paused. Update your card via **Change Plan** and reactivate — nothing is deleted.
## The 60-day guarantee [#the-60-day-guarantee]
If BlockMind isn't for you, contact us within your first 60 days at [contact@blockmind.app](mailto:contact@blockmind.app) and we'll refund you.
## Account basics [#account-basics]
* **Sign-in**: 6-digit email code or Google. No passwords.
* **Account deletion**: email [contact@blockmind.app](mailto:contact@blockmind.app) and we'll take care of it.
## Related [#related]
- [Usage](/docs/billing/usage)
- [Agent lifecycle](/docs/billing/agent-lifecycle)
- [Plans & pricing](/docs/overview/plans-and-pricing)
---
# Usage
Your agent's work is covered by a monthly usage allowance. You'll see it as a percentage — "72% left" — in the header pill and on your Usage page.
## Where to see it [#where-to-see-it]
* The **Usage pill** in the header/sidebar of your Elite dashboard.
* The **Usage page** at [blockmind.app/elite/credits](https://blockmind.app/elite/credits): your percentage remaining, your reset date, and a history of what used it.
## How it works [#how-it-works]
* **Resets monthly** on paid plans, on your billing cycle.
* **Morning Briefs are included** — they never draw from your usage.
* Ordinary chat is light. Heavier work draws more:
* Deep research runs (multi-source coin or narrative research)
* [Browser](/docs/capabilities/browser) sessions on live websites
* [KOL Analysis](/docs/capabilities/kol-analysis) in Discussion mode — the most usage-hungry of all, which is why the agent warns you before starting one
## Running low [#running-low]
* The Usage page shows your reset date — your allowance returns in full then.
* Morning Briefs keep arriving either way; they don't draw from usage.
## On the Free plan [#on-the-free-plan]
The Free plan has small classic-platform allowances — see [Plans & pricing](/docs/overview/plans-and-pricing) for what they are. The agent itself doesn't run on Free.
## Related [#related]
- [Billing & account](/docs/billing)
- [Plans & pricing](/docs/overview/plans-and-pricing)
---
# Browser
A lot of the information that matters in crypto never shows up in an API: a protocol's live dashboard, a governance vote, a docs page that changed yesterday. The Browser capability lets your agent open those pages itself, in its own cloud browser, and work with what it finds.
You don't operate the browser — you just ask. "Check the current parameters on this governance proposal." "Pull the numbers from this dashboard." Your agent opens the page, reads it, and brings the answer back into the conversation.
## What it can do [#what-it-can-do]
* **Read JS-heavy pages.** Modern dashboards and governance UIs render in the browser, not in raw HTML. Your agent sees the page the way you would.
* **Extract data.** Tables, metrics, vote tallies, parameter values — pulled off the page and into its answer.
* **Work through multi-step flows.** It can click through a site — open a page, apply a filter, drill into a detail view — when the information takes more than one step to reach.
* **Capture screenshots.** Visual proof of what a page showed at the time it looked.
* **Save pages as PDFs.** Useful when you want a whole page preserved, not just the numbers.
## What you see in chat [#what-you-see-in-chat]
Browser work is visible, not a black box. While your agent browses, the chat shows a **live session timeline** with screenshots, so you can watch what it's looking at as it works.
When the session ends, it publishes **artifacts** you can open directly from the conversation:
| Artifact | What it is |
| ---------- | ------------------------------------------------------- |
| Screenshot | An image of the page at a moment in the session |
| PDF | A saved copy of a full page |
| Data | Extracted values from the page, in a form you can reuse |
## When your agent reaches for it [#when-your-agent-reaches-for-it]
You rarely need to say "use the browser." Your agent picks it up when a question needs a live page — typical cases:
* A protocol dashboard whose numbers you want checked right now
* A governance proposal you want summarized before the vote closes
* A project site or docs page you want compared against what was there before
* Anything a normal web search returns as a link rather than an answer
If a plain web search or a data source can answer faster, it uses that instead — the browser is for pages that have to be *visited*.
## Usage [#usage]
Browser sessions are heavier than ordinary chat: keeping a live browser open and stepping through pages draws more of your monthly usage than a quick question does. That's usually fine for a handful of sessions, but worth knowing if you ask for a lot of deep page work. See [Usage](/docs/billing/usage) for how usage works and where to check what you have left.
Browser is a built-in capability, managed like any other from the My Agent page at [https://blockmind.app/elite/agent](https://blockmind.app/elite/agent).
Whatever your agent finds in the browser, it's research material — context for your decisions, not financial advice. Your agent will never tell you what to buy or sell.
## Related [#related]
- [Usage](/docs/billing/usage)
- [On-chain intelligence](/docs/capabilities/on-chain)
- [All capabilities](/docs/capabilities)
---
# Charts & dashboard
Two capabilities make your workspace visual: **Chart**, which renders interactive price charts inline in chat, and **Dashboard Layout**, which lets your agent rearrange your Elite home for you.
## Charts in chat [#charts-in-chat]
Ask for a chart the way you'd ask for anything else:
> "Show me a chart of SOL over the last three months."
Your agent renders an **interactive price chart directly in the conversation** — not a static image. You can look at price action right where the discussion is happening, alongside the analysis that goes with it.
### Overlays [#overlays]
Charts support moving-average overlays:
* **SMA** — simple moving average
* **EMA** — exponential moving average
Ask in plain language: "add the 50-day SMA" or "show ETH with a 20-day EMA." Overlays are context for reading a trend — your agent uses them to explain what it sees, never as buy or sell signals. Everything it tells you about a chart is research, not financial advice.
## The home dashboard [#the-home-dashboard]
Your Elite home is built from widgets — up to **11** of them, arranged the way you want. Open **Customize Dashboard** at [blockmind.app/elite/widgets](https://blockmind.app/elite/widgets) to choose and arrange them.
Available widgets include:
| Widget | What it shows |
| ----------------- | -------------------------------------------------- |
| Morning Brief | Your daily briefing — this is where it lands first |
| Agent Feed | What your agent has been doing |
| Chart | A live price chart on your home screen |
| Portfolio summary | Your holdings at a glance |
| Watchlist summary | The assets you're tracking |
| Notebook | Recent shared notes |
| Capabilities | What your agent can do right now |
There are more beyond these — the Customize Dashboard page shows the full set.
The chart widget uses live TradingView-powered charts, so the dashboard chart behaves like the real thing, not a snapshot.
## Or just ask [#or-just-ask]
You don't have to touch the customize page at all. The **Dashboard Layout** capability means your agent can edit the dashboard for you:
> "Add a chart widget for BTC to my dashboard."
>
> "Remove the watchlist summary and move the Notebook widget up."
Describe the layout you want and your agent makes it so. This is handy mid-conversation — if your agent surfaces something worth keeping an eye on, you can ask it to pin the relevant widget without leaving the chat.
## Related [#related]
- [Morning Brief](/docs/automations/morning-brief)
- [Watchlists](/docs/data/watchlists)
- [All capabilities](/docs/capabilities)
---
# Custom capabilities
Built-in capabilities cover the common ground. Custom capabilities cover *you*. With **Create Capability**, you teach your agent a workflow that's yours — the exact way you like something researched, summarized, or checked — and it becomes something the agent can repeat on request.
## How you create one [#how-you-create-one]
There's no form to fill in. You teach the workflow in conversation, the same way you'd explain it to a colleague:
> "When I ask for my stablecoin yield summary: group holdings by protocol, flag anything whose yield dropped more than a point since last time, and keep it under ten lines."
Your agent drafts the capability from what you described — the steps, the format, the details you care about — and shows you what it understood.
**Nothing is saved until you approve it.** The agent only stores the capability after your explicit go-ahead. If the draft misses something, correct it in the same conversation and approve the revised version instead.
## Using and managing them [#using-and-managing-them]
Once saved, reuse is just asking:
> "Run my stablecoin yield summary."
Your agent recognizes the workflow and executes it the way you taught it. Because the capability captures the format and the judgment calls, you get the same shape of output every time — no re-explaining.
To see what your agent has learned, open the **My Agent** page at [blockmind.app/elite/agent](https://blockmind.app/elite/agent). Your custom capabilities are listed alongside the built-in ones, and you can view or remove them there. You can also just tell your agent to drop one it no longer needs.
Creating custom capabilities is part of Pro — the same plan that includes your agent.
## What makes a good custom capability [#what-makes-a-good-custom-capability]
The best candidates are things you'd otherwise re-explain every time:
* A report format you like ("competitors first, then valuation, then risks — always in that order")
* A recurring summary with your own grouping and thresholds
* A checklist you run before acting on an idea
* A comparison you do often, in a specific structure
If a workflow is one-off, just ask for it directly — save capabilities for things you'll want again.
## Related ways to teach your agent [#related-ways-to-teach-your-agent]
Create Capability is one of three ways to make your agent work your way. Pick the one that matches what you're teaching:
| You want to teach... | Use |
| -------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------- |
| A repeatable workflow you run by asking | Create Capability (this page) |
| A check that should run on its own, every time the agent wakes | [Custom recurring checks](/docs/automations/monitoring-and-alerts) |
| Your own research framework for judging an idea | A framework note in your Notebook — it shows up as an expert in [Verdicts](/docs/capabilities/research-journey) |
However you teach it, the boundary is the same: your agent does research, not financial advice. A custom capability can shape *how* it analyzes, but it will never tell you what to buy or sell.
## Related [#related]
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
- [The research journey](/docs/capabilities/research-journey)
- [Notebook](/docs/memory/notebook)
---
# Capabilities
Capabilities are the things your agent can do. Each one is a self-contained ability — researching a token, reading on-chain flows, driving a cloud browser, drawing charts — that you can install, remove, and manage yourself. Your agent starts with **13 capabilities installed**, so it's useful from the first message.
This section covers each capability in depth. This page explains how to manage them.
## The My Agent page [#the-my-agent-page]
[My Agent](https://blockmind.app/elite/agent) is where you configure, monitor, and manage your AI analyst. It shows the full capability catalog, organized by category:
| Category | What lives here |
| ----------------------- | ----------------------------------------------------------------- |
| **Portfolio & Trading** | Capabilities grounded in your holdings, watchlists, and positions |
| **Research & Analysis** | Token, sector, and equities research; second opinions |
| **Market Intelligence** | On-chain flows, news, and market-wide context |
| **Utilities** | Charts, dashboard control, files, email, memory |
### Installing and removing [#installing-and-removing]
You can install and remove capabilities from the catalog at any time. Removing one doesn't delete anything your agent has learned — it just takes that ability out of rotation until you add it back.
### Capability statuses [#capability-statuses]
Some capabilities show a status badge:
* **Limited** — the capability works, but with reduced depth or coverage right now.
* **Needs setup** — it requires something from you first (for example, a connection) before your agent can use it.
* **Unavailable** — not currently usable on your workspace.
Some capabilities are marked **Pro capability** — they're part of the [Pro plan](/docs/overview/plans-and-pricing).
### Telegram lives here too [#telegram-lives-here-too]
The card for connecting your agent to Telegram is on the My Agent page. Telegram is rolling out gradually — if you don't see the connect card yet, it isn't available on your workspace quite yet. Once connected, you can message your analyst from your phone in the same conversation thread as the web. See [Telegram](/docs/channels/telegram) for details.
## Custom capabilities [#custom-capabilities]
Beyond the built-in catalog, you can teach your agent your own repeatable workflows in conversation — it saves them as custom capabilities with **Create Capability**, and you manage them from the same My Agent page. See [Custom capabilities](/docs/capabilities/custom).
## The capability catalog [#the-capability-catalog]
- [The research journey](/docs/capabilities/research-journey)
- [Use cases](/docs/capabilities/use-cases)
- [Research](/docs/capabilities/research)
- [Portfolio analysis](/docs/capabilities/portfolio-analysis)
- [On-chain intelligence](/docs/capabilities/on-chain)
- [KOL analysis](/docs/capabilities/kol-analysis)
- [Browser](/docs/capabilities/browser)
- [Charts & dashboard](/docs/capabilities/charts-and-dashboard)
- [Custom capabilities](/docs/capabilities/custom)
## Related [#related]
- [What your agent can do](/docs/overview/what-your-agent-can-do)
- [Memory & the Notebook](/docs/memory)
---
# KOL analysis
Different analysts see the same chart differently: a macro thinker asks about Fed direction and cycle position, a technician asks whether price reclaimed the level, a risk manager asks what breaks first. KOL Analysis lets you borrow those lenses — your agent analyzes a question in a specific analyst's voice and framework, so you can stress-test an idea against a way of thinking that isn't your own.
## The three modes [#the-three-modes]
You don't pick a mode from a menu — you describe what you want, and your agent picks the right depth.
### Q\&A — one analyst, one answer [#qa--one-analyst-one-answer]
Ask a question through a single analyst's framework. Fast and focused: "How would a macro-first analyst read this market?" You get one coherent perspective, applied to your question.
### Synthesis — two frameworks, one view [#synthesis--two-frameworks-one-view]
Have two frameworks look at the same question and merge into a single answer: where they agree, where they diverge, and what the combined read is. Useful when your idea looks great on the chart but you're unsure about the macro — or the other way around.
### Discussion — full cross-critique [#discussion--full-cross-critique]
The deep mode: a full discussion between experts, where each framework critiques the others' reasoning rather than just answering in parallel. Positions get challenged, weak arguments get exposed, and you see *why* the perspectives disagree, not just that they do.
Discussion mode is genuinely heavy: it takes **5–20 minutes** to run and uses noticeably more of your monthly [usage](/docs/billing/usage) than a normal question. Your agent warns you before starting one, so you won't kick it off by accident.
Save Discussion mode for decisions that deserve it — a position you're seriously weighing, a thesis you want properly attacked. For quick perspective checks, Q\&A or Synthesis gets you most of the value at a fraction of the cost and time.
## How to invoke it [#how-to-invoke-it]
Just ask in chat, naming the perspective you want:
* "How would a pure technical analyst read my ETH position right now?"
* "Give me a macro take and a chart take on this, and synthesize them."
* "Run a full discussion on whether this narrative has legs — I want the frameworks to argue."
## KOL Analysis and Verdicts [#kol-analysis-and-verdicts]
KOL Analysis and Verdicts are related but sit in different places. **Verdicts** — the fourth stop in the [Explore → Analyze → Verdicts → Track journey](/docs/capabilities/research-journey) — are structured second opinions on a saved analysis, delivered by built-in experts with distinct frameworks (and any custom frameworks you've saved in your Notebook). **KOL Analysis** is the conversational version: the same idea of framework-based perspectives, available on any question in chat, at whatever depth you choose — no saved analysis required.
Rough guide: working through a saved idea inside the journey → Verdicts. Ad-hoc "how would X see this?" in the middle of a conversation → KOL Analysis.
Every mode — including a full Discussion — is research, not financial advice. The experts argue frameworks and evidence; none of them will ever tell you what to buy or sell, and your agent can never touch your funds.
## Related [#related]
- [Explore → Analyze → Verdicts → Track](/docs/capabilities/research-journey)
- [Research](/docs/capabilities/research)
- [Usage](/docs/billing/usage)
---
# On-chain intelligence
Price tells you what happened; on-chain data tells you who did it and where the money actually sits. The On-Chain Intelligence capability gives your agent both sides of that picture, and you get at it the way you get at everything else — by asking in plain language.
## Money movement — powered by Nansen [#money-movement--powered-by-nansen]
Your agent can look at who is moving what:
* **Whale and smart-money flows.** What large and historically successful wallets are accumulating or distributing, and whether flows into a token are broad or concentrated in a few hands.
* **Token holder analysis.** How a token's holder base is structured — concentration, growth, and how it's changing over time.
* **Wallet profiling.** Point your agent at a public address and ask what kind of wallet it looks like and what it's been doing.
## Protocol fundamentals — powered by DeFiLlama [#protocol-fundamentals--powered-by-defillama]
The other half is protocol health — the on-chain equivalent of fundamentals:
* **TVL** — total value locked in a protocol or chain, and how it's trending.
* **Yields** — where returns are on offer, and how they compare.
* **Fees and revenue** — whether a protocol actually earns anything.
* **DEX volumes** — where trading activity is happening on-chain.
## Example asks [#example-asks]
* "What are whales doing in ETH this week?"
* "Profile this wallet: 0x… — what's it been accumulating?"
* "How concentrated is this token's holder base?"
* "Is smart money moving into this narrative or out of it?"
* "Compare TVL and fee revenue across the top lending protocols."
* "Has this protocol's revenue kept up with its token price?"
* "Where are DEX volumes growing right now?"
## Combining it with everything else [#combining-it-with-everything-else]
On-chain data rarely answers a question alone — the value is in the cross-check. Ask your agent to run [research](/docs/capabilities/research) on a token *and* check what holders and smart money are doing, and you get one answer where the narrative and the on-chain reality either line up or visibly don't. Since your agent also knows [your portfolio](/docs/capabilities/portfolio-analysis), "check the holder structure of everything I own" is a fair ask.
On-chain intelligence is research, not financial advice. Smart-money flows describe what other wallets did — your agent will never turn that into a buy or sell instruction, and it can never touch your funds.
## Related [#related]
- [Research](/docs/capabilities/research)
- [Portfolio analysis](/docs/capabilities/portfolio-analysis)
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
---
# Portfolio analysis
Portfolio analysis isn't a feature you switch on — it's the context your agent carries everywhere. Once your portfolio is connected, every relevant answer is grounded in what you actually hold: research weighs your positions, news is filtered through your exposure, and "the market is down" becomes "here's what that means for *you*."
This page assumes your portfolio is connected. If it isn't yet, start with [Connect your portfolio](/docs/getting-started/connect-your-portfolio), including the exchange-key scope details.
## What to ask [#what-to-ask]
Anything you'd ask a human analyst who has your statement open in front of them:
### Holdings and allocation [#holdings-and-allocation]
* "Break down my holdings by chain and by asset."
* "What's my largest position, and how has it changed this month?"
* "How much of my portfolio is stablecoins right now?"
### Exposure [#exposure]
* "What's my real exposure to ETH — including LSTs and L2 tokens?"
* "Which narratives am I concentrated in?"
* "If BTC drops 15%, which of my positions are most correlated?"
### Performance [#performance]
* "Why is my portfolio down today?"
* "Which positions drove my performance this week?"
* "Compare my portfolio's performance to just holding BTC."
### Position changes [#position-changes]
* "What changed in my portfolio since we last talked?"
* "Have my exchange balances changed since last week?"
Your agent answers from your live connected data — no exports, no copy-pasting balances into chat.
## Where portfolio context shows up [#where-portfolio-context-shows-up]
Your portfolio isn't just there when you ask about it directly. It feeds:
* **Every research answer.** Ask about a token you hold and your agent knows it — the answer accounts for your position, not a hypothetical one.
* **The [Morning Brief](/docs/automations/morning-brief).** Overnight moves are explained in terms of your holdings: what moved, why, and what it means for you.
* **[Monitoring passes](/docs/automations/monitoring-and-alerts).** Between briefs, your agent scans your portfolio for large moves and flags what crosses your thresholds.
## Multiple portfolios [#multiple-portfolios]
You can connect several wallets and exchange accounts, and your agent sees across all of them. Ask about the whole picture ("total exposure across everything") or a slice ("just my cold wallet"). The [Portfolios page](/docs/data/portfolios) gives you the same multi-portfolio overview with drill-down, outside of chat.
## What your agent can and can't do here [#what-your-agent-can-and-cant-do-here]
Your agent can see your positions and analyze them. It has no tools to trade, withdraw, or transfer, and it will never tell you what to buy or sell. BlockMind uses connections only to read balances and positions; see [Connect your portfolio](/docs/getting-started/connect-your-portfolio) for the Firi key-scope exception. It does the analysis; the decisions are yours.
## Related [#related]
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
- [Morning Brief](/docs/automations/morning-brief)
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
---
# The research journey
Most research tools give you a blank search box. BlockMind gives you a pipeline: **Explore → Analyze → Verdicts → Track**. Each stage takes the output of the one before it, and every meaningful output gets a stable NB-number in your [Notebook](/docs/memory/notebook) — so an idea you found in June is still there, with its analysis and second opinions attached, when you revisit it in October.
You don't have to use all four stages every time. But together they turn "I saw something interesting" into a documented thesis you can actually follow.
## Explore — find ideas [#explore--find-ideas]
[Explore](https://blockmind.app/elite/explore) runs guided market scans. Instead of writing a prompt, you pick three things:
| Dimension | Options |
| --------------- | ---------------------------- |
| **Asset class** | Crypto, or stocks & equities |
| **Timeframe** | Near, mid, or long term |
| **Market cap** | Micro, medium, or large |
Your agent scans for candidates that fit and returns them with context. Anything worth keeping, you **save as an idea** — it gets an NB-number and appears in your Notebook. Saved ideas can be analyzed deeper on the Analyze page.
## Analyze — go deep on one idea [#analyze--go-deep-on-one-idea]
[Analyze](https://blockmind.app/elite/analyze) is one button on a saved idea. Your agent produces a structured analysis covering:
* **Conviction** — how strong the case is, and what it rests on
* **Levels** — where the idea is validated or invalidated
* **Competitors** — who else is playing in the same space
* **Valuation** — how the price relates to what's underneath it
Saved analyses get their own NB-numbers, linked to the idea they came from.
## Verdicts — get second opinions [#verdicts--get-second-opinions]
Verdicts is where an analysis gets stress-tested. Four built-in experts each read it through a distinct framework and give their judgement:
| Expert | Framework |
| ------------------- | --------------------------------------------------------------------------------------------------- |
| **Oskar Solberg** | "I check the macro first: Fed direction, stablecoin flows, cycle position." |
| **Louis Solberg** | "I check the chart: did price reclaim the level, is momentum improving, is the invalidation clean?" |
| **Dr. Marcus Voss** | "I check the risk: leverage in the system, crowded positions, single points of failure." |
| **Dorian Vale** | "I check whether the trade is ready… reward at least 2× the risk." |
Because the frameworks are different, they disagree in useful ways — a chart that looks ready to Louis might still be too crowded for Dr. Voss. That tension is the point.
### Your own frameworks as experts [#your-own-frameworks-as-experts]
The four built-ins are a starting bench, not the whole team. Any **framework note** saved in your Notebook appears as an expert on the Verdicts page — so your own decision process weighs in alongside theirs. Ask your agent to save your decision process as a framework note; see [The Notebook](/docs/memory/notebook) for how framework notes work.
Verdicts are research judgements, never trade instructions. Your agent does research, not financial advice — it will never tell you what to buy or sell, and it can never touch your funds. The decision is always yours.
## Track — follow what you own and watch [#track--follow-what-you-own-and-watch]
[Track](https://blockmind.app/elite/track) is the ongoing stage — where ideas you've acted on (or are still watching) get followed over time:
* **Portfolio snapshot** — your holdings at a glance
* **Position snapshots** — per-position price trends, news, events, and catalysts
* **Brief context** — what your [Morning Brief](/docs/automations/morning-brief) has been saying about your positions
* **Track your sources** — follow the analysts and sources you trust
## How it fits together [#how-it-fits-together]
A worked example:
1. **Explore**: scan crypto, mid-term, micro cap. Three candidates come back; you save one as an idea.
2. **Analyze**: one button. Conviction, levels, competitors, valuation — saved with an NB-number.
3. **Verdicts**: four experts (plus your own framework, if you've saved one) weigh in. Two are constructive, one flags crowding.
4. **Track**: you take a position — or don't — and Track keeps the news, catalysts, and price action for it in one place.
Every step lives in your Notebook, so "why did I like this?" always has an answer.
## Related [#related]
- [The Notebook](/docs/memory/notebook)
- [KOL analysis](/docs/capabilities/kol-analysis)
- [Custom capabilities](/docs/capabilities/custom)
---
# Research
Research is what your agent does most. Four built-in capabilities cover the core of it, and they work together: ask one question and your agent decides which to draw on, combining market data, the open web, and X/Twitter into one answer.
## The research capabilities [#the-research-capabilities]
| Capability | What it covers |
| ---------------------- | ----------------------------------------------------------------------------------------- |
| **Coin Research** | Structured research on a specific token or protocol — market data, the web, and X/Twitter |
| **Research Narrative** | Sectors, themes, and capital flows — the bigger picture around an asset |
| **Equities Research** | Public stocks — quotes and fundamentals, plus ETFs, indexes, and commodities data |
| **News** | What's being reported, and what it means for what you hold |
### Coin Research [#coin-research]
Ask about a token or protocol and your agent runs structured research on it: current market data, what's being written about it on the web, and what the conversation looks like on X/Twitter. Good for "should I look deeper at this?" moments — a coin someone mentioned, a protocol in your watchlist, a position you're reconsidering.
### Research Narrative [#research-narrative]
Where Coin Research goes deep on one asset, Research Narrative goes wide: sectors, themes, and where capital is flowing. Ask about a narrative ("what's happening in DeFi lending?") or ask which narratives are gaining momentum right now.
### Equities Research [#equities-research]
Your agent isn't crypto-only. It can pull quotes and fundamentals on public stocks, and covers ETFs, indexes, and commodities data too — useful when your thesis crosses asset classes ("how does this miner compare to holding BTC directly?").
### News [#news]
Your agent reads the news so you can ask about it: what happened, why it matters, and — because it knows your portfolio — what it means for you specifically.
## How to run research [#how-to-run-research]
**Just ask in chat.** There's no mode to switch into — describe what you want to know and your agent picks the right capabilities:
* "Research this token before I decide anything."
* "What's the narrative around real-world assets right now?"
* "Pull up fundamentals on this stock and compare it to the sector."
* "Any news today that affects my holdings?"
**Or use the guided journey.** The [Explore and Analyze pages](/docs/capabilities/research-journey) wrap these same capabilities in a structured flow: scan for ideas on Explore, then run a one-button structured analysis on Analyze — conviction, levels, competitors, valuation.
## Depth and usage [#depth-and-usage]
Research can be quick or deep, and you control which. A fast take on a coin is light; a thorough multi-source deep dive draws more of your monthly [usage](/docs/billing/usage). If you want the deep version, say so ("go deep on this") — and if you're just triaging, ask for a quick read first.
## Saving what you find [#saving-what-you-find]
Good research shouldn't evaporate when the chat ends. Ask your agent to save a result to your [Notebook](/docs/memory/notebook) and it becomes a durable note with a stable NB-number you can both reference later — "compare this to what we found in NB-128."
Your agent does research, not financial advice. It will never tell you what to buy or sell — it gives you the analysis, and the decision stays yours.
## Related [#related]
- [Explore → Analyze → Verdicts → Track](/docs/capabilities/research-journey)
- [On-chain intelligence](/docs/capabilities/on-chain)
- [Usage](/docs/billing/usage)
---
# Use cases
The hardest part of working with an agent is the blank box: you know it can do a lot, but you're not sure what to type. [Use cases](https://blockmind.app/elite/use-cases) fixes that with **36 curated prompts across 6 categories** — buttons, not blank boxes.
Each card carries a full, expert-written prompt behind it. Click one and it launches a fresh session with your agent already briefed on exactly what to do and how to do it well. You get the output of a carefully engineered prompt without writing it.
## Today's use case [#todays-use-case]
The page rotates a featured **"Today's use case"** — a different prompt highlighted each day. It's a low-effort way to discover parts of the catalog you haven't tried.
## The six categories [#the-six-categories]
### Today's edge [#todays-edge]
What's moving right now and why — the prompts you run when you sit down in the morning or after a volatile day. Think of it as pulling the day's signal out of the noise, grounded in your holdings.
### Find the next play [#find-the-next-play]
Idea generation: scanning for candidates that fit a theme, a market-cap range, or an emerging narrative before they're obvious. These pair naturally with saving results as ideas in the [research journey](/docs/capabilities/research-journey).
### Your portfolio [#your-portfolio]
Prompts aimed squarely at what you hold — health checks, concentration and exposure reviews, "explain what happened to my positions" sessions. Everything is grounded in your live connected [portfolio](/docs/data/portfolios).
### Smart money & on-chain [#smart-money--on-chain]
What large holders and smart-money wallets are actually doing, and what the chain data says beneath the price action. Backed by the [On-chain intelligence](/docs/capabilities/on-chain) capability.
### Second opinions [#second-opinions]
Stress-test a thesis before you commit to it — get an idea challenged from angles you didn't think of, in the spirit of the [Verdicts](/docs/capabilities/research-journey) stage.
### On autopilot [#on-autopilot]
Set-and-forget: prompts that put your agent to work in the background — recurring checks and standing watch. See [Monitoring & alerts](/docs/automations/monitoring-and-alerts) for how the background side works.
## When to use it [#when-to-use-it]
* **You're new** and want to see what a well-prompted session looks like before writing your own.
* **You're busy** and want a one-click version of something you'd otherwise type out.
* **You're stuck** — the catalog doubles as a menu of what your agent is good at.
Once a session launches, it's a normal conversation: follow up, redirect, ask for more depth. The card is just the starting push.
Everything these prompts produce is research, not financial advice. Your agent will never tell you what to buy or sell, and it can never touch your funds.
## Related [#related]
- [The research journey](/docs/capabilities/research-journey)
- [Your first conversation](/docs/getting-started/first-conversation)
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
---
# Email
Your agent has its own email inbox. That makes email a real two-way channel: it can write to you, you can write back, and the conversation continues — no app required.
## What your agent emails you [#what-your-agent-emails-you]
* **The Morning Brief.** Your daily briefing lands in your inbox as well as on your dashboard — see [Morning Brief](/docs/automations/morning-brief).
* **Alerts.** Email is one of the delivery channels for alerts you've set, alongside in-app and Telegram.
* **Anything you ask for.** "Email me that analysis" works — including files from its workspace, like a report or a data file it produced.
As everywhere else, what it sends is research and context — it will never email you instructions to buy or sell.
## Reply to continue the conversation [#reply-to-continue-the-conversation]
Replying to any email from your agent **wakes it up** and continues the conversation right there in your email thread. Ask a follow-up question about the Morning Brief, push back on a point in a report, or hand it a new task — the reply reaches your agent like a chat message would, and the conversation continues from there.
## Ask it to handle email for you [#ask-it-to-handle-email-for-you]
Beyond emailing *you*, your agent can work with email on request:
* **Send** an email — for example, mail a summary or a file to yourself for later.
* **Reply** to an email in its inbox.
* **Forward** something it received, or something from its workspace.
You direct all of this in plain language from any channel: "forward me that PDF", "reply and ask for the details".
Sending is queued, not instant. Emails from your agent usually arrive promptly, but don't rely on second-level delivery — for anything urgent, the web chat or Telegram is faster.
## Ideas for using it [#ideas-for-using-it]
* **Push back on the brief.** Reply to the morning email with "why does this matter for my ETH position?" and continue the conversation right there.
* **Traveling light?** Ask for the Morning Brief and any follow-ups by mail, and run the whole exchange from your inbox.
* **Paper trail.** Ask it to email you a finished analysis so you have a copy where you keep everything else.
## Related [#related]
- [Morning Brief](/docs/automations/morning-brief)
- [Telegram companion](/docs/channels/telegram)
- [Web chat](/docs/channels/web-chat)
---
# Telegram companion
The Telegram companion puts your agent in your pocket. It's not a separate bot with its own personality and its own history — it's the **same conversation** you have on the web, continued from your phone.
## Connect [#connect]
1. Open [My Agent](https://blockmind.app/elite/agent) or [Settings](https://blockmind.app/elite/settings) and look for the Telegram connect card.
2. Follow the connect link to open Telegram and link your account.
Connect links expire after **60 minutes**. If yours has lapsed by the time you use it, just generate a fresh one from the same place.
Each account gets **one private Telegram chat** — it's a direct line between you and your agent, not a group or a shared channel.
Telegram is still rolling out. If you don't see the connect card on My Agent or Settings yet, it isn't available on your workspace yet — it's on its way.
## What you can send [#what-you-can-send]
Everything you'd expect from a Telegram chat works:
* **Text** — ask anything, exactly as you would on the web.
* **Voice notes** — speak your question; it's transcribed and answered. Your agent can reply with voice too.
* **Photos and screenshots** — snap a chart or paste a screenshot and ask about it.
* **Document files** — send a file for your agent to read.
## What arrives there [#what-arrives-there]
Beyond replies to your messages, your agent can reach you on Telegram proactively:
* **Alerts** you've set — price levels, daily moves, ratios — can be delivered there alongside in-app and email. See [Monitoring and alerts](/docs/automations/monitoring-and-alerts).
The daily **Morning Brief** is published to the BlockMind dashboard and sent by email. It is not delivered through Telegram.
Anything your agent surfaces on Telegram is research and context, never an instruction to buy or sell.
## One conversation, wherever you are [#one-conversation-wherever-you-are]
Because Telegram and the web share one thread, you can start something on your phone — "look into this token, I'll check when I'm home" — and pick up the full result on the Elite dashboard later. The complete history always lives in the [web app](/docs/channels/web-chat); Telegram is a window into it, not a copy of it.
## Disconnect and reconnect [#disconnect-and-reconnect]
* Send **/stop** in the Telegram chat to disconnect.
* **Reconnect** any time from [Settings](https://blockmind.app/elite/settings).
Disconnecting doesn't delete anything — your conversation history stays in the web app either way.
## Related [#related]
- [Web chat](/docs/channels/web-chat)
- [Email](/docs/channels/email)
- [Monitoring and alerts](/docs/automations/monitoring-and-alerts)
---
# Web chat
The web chat on your Elite dashboard is the main way to talk to your agent. It's where every conversation lives in full — whatever you send from [Telegram](/docs/channels/telegram) lands in the same thread here, and replying to your agent's [emails](/docs/channels/email) continues the same conversation.
## Sessions [#sessions]
Conversations are organized into sessions:
* **Date-grouped tabs** keep recent sessions one click away.
* **A searchable all-sessions list** lets you find any past conversation.
* **Rename** a session to something you'll recognize later.
* **New chat** starts a fresh session whenever you want a clean slate.
Your agent's memory spans sessions — starting a new chat doesn't make it forget what it knows about you. It just gives the conversation a fresh thread.
## While your agent works [#while-your-agent-works]
When you send a message, the response streams in live. Two things worth knowing:
* **Status labels show what it's doing.** While your agent works, you'll see per-tool labels — thinking, browsing, analyzing — so a longer task never feels like a black box.
* **Streams survive reloads.** If you refresh the page or your connection drops, you'll see "Reconnecting…" and the response picks up where it left off. You can close the tab mid-answer and come back to the finished result.
There's no stop button. Once your agent starts a response, it runs to completion — if it's heading the wrong way, send a follow-up message and it will redirect from there.
## Attachments [#attachments]
You can attach files to any message — up to **5 files per message, 4 MB each**. Supported formats:
| Type | Formats |
| --------- | ------------------------------- |
| Documents | PDF, DOCX, Markdown, plain text |
| Data | XLSX, CSV, JSON |
| Images | PNG, JPEG, GIF, WebP |
Screenshots are the fast path: copy one and **paste it straight into the composer**. "What's this chart telling me?" with a pasted screenshot is one of the most useful things you can send.
## What your agent can send back [#what-your-agent-can-send-back]
Replies aren't limited to text:
* **Interactive charts** rendered inline — ask for a price chart and you get one you can explore, with SMA/EMA overlays on request.
* **Browser artifacts** — when your agent uses its [cloud browser](/docs/capabilities/browser), you can follow a live timeline with screenshots, and open the screenshots, PDFs, and data files it publishes.
* **Emoji reactions** to your messages.
* **Voice replies** you can play — ask for a spoken answer and pick a male or female voice if you have a preference.
Voice goes one way for now: your agent can speak to you, but the web chat doesn't take voice input. If you prefer talking, voice notes work on [Telegram](/docs/channels/telegram).
## A note on what you'll get [#a-note-on-what-youll-get]
Everything your agent tells you in chat is research — analysis, data, and context to inform your own decisions. It will never tell you what to buy or sell, and it can never touch your funds.
## Related [#related]
- [Telegram companion](/docs/channels/telegram)
- [Your first conversation](/docs/getting-started/first-conversation)
- [Charts and your dashboard](/docs/capabilities/charts-and-dashboard)
---
# DeepDive reports
DeepDive reports are free AI research reports on the classic platform. Pick a coin, and BlockMind researches it and writes up what it finds — no agent required.
## Get a report without an account [#get-a-report-without-an-account]
The public flow lives at [blockmind.app/deepdive](https://blockmind.app/deepdive):
1. Enter the coin you want researched.
2. Enter your email address.
3. The report generates asynchronously — typically within minutes — and you get a link to it by email.
You don't need to sign up, and you don't need to wait on the page.
The public flow is limited to **5 reports per day per email address**.
## Request analyses when signed in [#request-analyses-when-signed-in]
Signed in, you can request analyses directly from your holding and watchlist cards — the report is generated for the coin on that card and lands in your Inbox under Reports.
| Plan | Analysis requests |
| ---------- | ---------------------- |
| Free | 3 per month |
| Paid plans | 30 per rolling 30 days |
## What a report covers [#what-a-report-covers]
Each report is a structured first pass on one coin — what the project is, how it's trading, and what the data says about it. The goal is a fast, honest read: enough to decide whether a coin deserves more of your attention, delivered in minutes instead of an afternoon of tabs.
## DeepDive vs your agent [#deepdive-vs-your-agent]
A DeepDive is a one-shot report on one coin. If you're on Pro, your agent's [research](/docs/capabilities/research) goes further in three ways:
* **Deeper** — it can pull in more sources, follow threads, and take multiple passes at a question.
* **Interactive** — a report ends; a conversation doesn't. Ask follow-ups, challenge conclusions, zoom in on the part you care about.
* **Grounded in your portfolio** — the agent knows what you hold, so its research connects to your actual positions instead of standing alone.
Think of DeepDives as the classic platform's research, and the agent as research that knows you.
## A note on what reports are [#a-note-on-what-reports-are]
DeepDive reports are research, not financial advice. They will never tell you what to buy or sell — they lay out what the data says so you can make your own call. Verify anything important before acting on it.
## Related [#related]
- [The classic platform](/docs/classic/overview)
- [Agent research](/docs/capabilities/research)
- [Plans & pricing](/docs/overview/plans-and-pricing)
---
# Market indicators
BlockMind tracks four market-wide indicators at [blockmind.app/market/indicators](https://blockmind.app/market/indicators). Together they answer one question from four angles: *what mood is the market in right now?*
All four indicators update **once daily**. They are not real-time gauges — treat them as a daily reading of market temperature, not a live feed.
## The four indicators [#the-four-indicators]
### Fear & Greed Index [#fear--greed-index]
The classic crypto sentiment gauge, sourced from Alternative.me. It scores overall market sentiment from 0 (extreme fear) to 100 (extreme greed). Low readings mean the market is anxious; high readings mean it's euphoric. Historically, extremes in either direction have been more interesting than the middle.
### BTC Dominance [#btc-dominance]
Bitcoin's share of the total crypto market cap, via LunarCrush data. Rising dominance usually means capital is favoring Bitcoin over altcoins; falling dominance often accompanies risk-on rotation into alts.
### Altcoin Season Index [#altcoin-season-index]
A BlockMind in-house indicator: it measures how many of the top 50 altcoins outperformed Bitcoin over the last 90 days, scored 0 to 100. High readings suggest altcoins are broadly leading; low readings suggest Bitcoin is doing the heavy lifting.
### Contrarian Leverage Index [#contrarian-leverage-index]
A BlockMind in-house sentiment gauge built from derivatives positioning data via Coinglass — open interest, long/short positioning, and funding rates — scored 0 to 100.
The Contrarian Leverage Index is a sentiment score, not a signal generator. It does not emit buy or sell signals. An extreme reading describes how traders are positioned — it is never an instruction to act.
## Free access and Pro context [#free-access-and-pro-context]
All four indicators are included on the free [classic platform](/docs/classic/overview).
On Pro, your agent can read the same indicator data and relate today's readings to your holdings, research, and Morning Brief.
## On each indicator page [#on-each-indicator-page]
Every indicator has its own page with:
* **History** — how the reading has moved over time, so you can see today's value in context.
* **AI chat** — ask questions about that indicator directly on the page: what it measures, why it moved, what similar readings looked like before.
## How to use them sensibly [#how-to-use-them-sensibly]
* **Read them as context, not triggers.** An indicator tells you the market's mood; it doesn't tell you what to do about it.
* **Expect lag.** All four update once daily and summarize what has already happened. They confirm conditions more often than they predict turns.
* **Extremes can persist.** "Extreme greed" can stay extreme for weeks in a strong trend, and fear can grind on through a long drawdown. An extreme reading is not, by itself, a reversal.
* **Combine them.** BTC Dominance plus the Altcoin Season Index together say more about rotation than either alone; Fear & Greed plus the Contrarian Leverage Index together say more about crowding than either alone.
These indicators are research inputs, not financial advice. Neither the indicators nor your agent will ever tell you what to buy or sell — the decision is always yours.
## Related [#related]
- [The classic platform](/docs/classic/overview)
- [DeepDive reports](/docs/classic/deepdive-reports)
- [Plans & pricing](/docs/overview/plans-and-pricing)
---
# The classic platform
The classic platform is the original BlockMind dashboard — and it's the free tier. It's still maintained, still improving, and not going anywhere. If you want portfolio tracking, watchlists, market indicators, and a taste of AI research without a personal agent, this is where you live.
Everyone gets it: sign up for free and it's yours. Pro users keep every classic feature too, alongside their agent.
## What's included, free [#whats-included-free]
### Portfolio home [#portfolio-home]
Your portfolio lives at [blockmind.app/portfolio](https://blockmind.app/portfolio): connect wallets and exchanges for balance-only use or build portfolios manually, and see your holdings in one place. Connection steps and exchange-key scope details are covered in [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
### Watchlists [#watchlists]
Track coins you don't hold yet. Watchlists sit next to your portfolios and count toward the same collection limit (see the table below).
### Live charts [#live-charts]
Charts across the platform are live and TradingView-powered, so you get proper charting rather than a static sparkline.
### Market Signals [#market-signals]
[Market Signals](https://blockmind.app/market/signals) is your 24/7 market scanner — a continuously updated feed of what's moving across the market. It's currently free for all users.
### Market indicators [#market-indicators]
All four [market indicators](/docs/classic/market-indicators) are free: the Fear & Greed Index, BTC Dominance, the Altcoin Season Index, and the Contrarian Leverage Index. Each has its own page with history and an AI chat about that indicator.
### Inbox [#inbox]
Your Inbox collects two things: **Updates** (what's changed that concerns you) and **Reports** (the research reports you've requested).
### DeepDive reports and the Coin Health Check [#deepdive-reports-and-the-coin-health-check]
Anyone can request a free AI research report on a coin — see [DeepDive reports](/docs/classic/deepdive-reports). There's also the [Coin Health Check](https://blockmind.app/check), a quick public look at any coin's condition.
## Free plan limits [#free-plan-limits]
| Feature | Free limit |
| ---------------------------------------------- | ----------- |
| AI chat messages | 5 per day |
| Analysis requests | 3 per month |
| Collections (portfolios + watchlists combined) | 5 |
| Market indicators | All 4 |
## What the classic platform does not include [#what-the-classic-platform-does-not-include]
The free tier is the platform without the agent. You don't get:
* **Your personal agent** — the named analyst with its own workspace, memory, and capabilities.
* **Morning Briefs** — the scheduled daily briefing grounded in your holdings, set around 8:00 by default.
* **Monitoring and alerts** — the agent checking your portfolio and the market throughout the day.
Those come with Pro. See [Plans & pricing](/docs/overview/plans-and-pricing) for what Pro includes and what it costs.
Upgrading doesn't take the classic platform away. Pro removes the free limits — chat, analysis requests (up to 30 per rolling 30 days), and collections — and adds the agent on top.
Everything on the classic platform, like everything in BlockMind, is research — not financial advice. The AI will never tell you what to buy or sell.
## Related [#related]
- [Market indicators](/docs/classic/market-indicators)
- [DeepDive reports](/docs/classic/deepdive-reports)
- [Plans & pricing](/docs/overview/plans-and-pricing)
---
# Portfolios
The Portfolios page at [blockmind.app/elite/portfolios](https://blockmind.app/elite/portfolios) is where everything you hold comes together: a multi-portfolio overview at the top, with drill-down into any single portfolio for its holdings and detail.
Portfolios aren't just a tracking view. They're the context your agent works from — every answer, brief, and monitoring pass starts with what you actually hold.
## Ways to build a portfolio [#ways-to-build-a-portfolio]
| Method | Best for |
| ----------------------- | --------------------------------------------------------------------------------------- |
| **Connect a wallet** | On-chain holdings — paste any public address or connect a wallet |
| **Connect an exchange** | Balances held on an exchange, using read-only key scopes where the provider offers them |
| **CSV import** | Bringing history over from another tracker or a spreadsheet |
| **Manual entry** | Positions you want to track by hand, or assets held elsewhere |
The step-by-step mechanics for each — supported chains, supported exchanges, CSV format — live in [Connect your portfolio](/docs/getting-started/connect-your-portfolio). BlockMind uses connections only to read balances and positions; the Firi credential exception is explained there.
## Wallet-linked vs manual portfolios [#wallet-linked-vs-manual-portfolios]
The two kinds behave differently, and it helps to know which you're looking at:
* **Wallet-linked portfolios** mirror the chain. Holdings come from what the address actually holds, so you can't freely edit wallet-derived positions — the chain is the source of truth. If a balance looks stale, re-scan the wallet to refresh it.
* **Manual portfolios** are yours to shape. Add, edit, and remove positions freely — useful for tracking a strategy on paper or holdings BlockMind can't see directly.
Exchange-connected portfolios sit closer to the wallet-linked model: balances reflect what the exchange reports.
## Spam and dust filtering [#spam-and-dust-filtering]
Wallets attract junk — airdropped spam tokens and dust balances that would otherwise clutter your holdings and distort your portfolio value. BlockMind filters suspicious tokens by default.
If something you actually hold got caught by the filter, the suspicious-tokens toggle reveals everything that's been filtered. If a legitimate token is missing entirely, a re-scan of the wallet is the first thing to try.
## Privacy [#privacy]
Portfolios are private by default. Your holdings live in your account and your agent's workspace — they're not shared with other users. See [Trust & security](/docs/security) for the full picture of what's stored and how connections stay read-only.
## How your agent uses your portfolios [#how-your-agent-uses-your-portfolios]
Your portfolios are the grounding layer for almost everything the agent does:
* **Answers in chat.** "Why is my portfolio down today?" is answered against your live holdings, not a generic market summary.
* **Morning Brief.** The "what matters to you" section is built from your positions. See [Morning Brief](/docs/automations/morning-brief).
* **Monitoring.** The agent's background passes scan your portfolio for large moves and flag them. See [Monitoring & alerts](/docs/automations/monitoring-and-alerts).
* **Analysis on demand.** Ask for a full portfolio review — allocation, concentration, risk — via [Portfolio analysis](/docs/capabilities/portfolio-analysis).
What your agent does with this data is research, not financial advice — it will never tell you what to buy or sell, and it can never touch your funds.
## Limits on the free plan [#limits-on-the-free-plan]
On the free plan, portfolios and [watchlists](/docs/data/watchlists) count toward a combined limit of 5 collections. See [Limits](/docs/reference/limits) for the complete table.
## Related [#related]
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
- [Portfolio analysis](/docs/capabilities/portfolio-analysis)
- [Watchlists](/docs/data/watchlists)
---
# Watchlists
The Watchlists page at [blockmind.app/elite/watchlists](https://blockmind.app/elite/watchlists) is where assets you don't hold yet — or ideas you're still forming — live. On the Elite dashboard, a watchlist is more than a price list: it's a thesis tool, with entry targets, target states, and strategy settings per list.
## Entry targets and "Near target" [#entry-targets-and-near-target]
For each asset on a watchlist you can set an **entry target** — the price at which the idea becomes interesting to you. From then on:
* The watchlist shows how each asset sits relative to its target, and flags assets in a **"Near target"** state when price approaches the level you set.
* When a target is hit, you get a **target alert** in your Inbox — so a level you set weeks ago doesn't slip past unnoticed.
Alerts are evaluated during your agent's monitoring passes, so expect them within hours rather than tick-by-tick. More on timing and delivery in [Monitoring & alerts](/docs/automations/monitoring-and-alerts).
## Strategy settings per watchlist [#strategy-settings-per-watchlist]
Each watchlist has its own strategy settings, so different lists can mean different things. A "long-term accumulation" list and a "short-term momentum" list don't have to be judged the same way — set the strategy per list and both the page and your agent read your assets in that context.
## Your agent reads — and writes — your watchlists [#your-agent-reads--and-writes--your-watchlists]
Watchlists are one of the places where your agent works alongside you, not just for you. It can read them for context, and it can edit them when you ask. Try:
* **"Add the top three names from that research to my DeFi watchlist"** — promote candidates straight from a [research](/docs/capabilities/research) session.
* **"Go through my watchlists and suggest what to prune"** — clear out stale ideas whose thesis has expired.
* **"Why is SOL showing near target?"** — get the context behind a target state: what moved, and whether your original thesis still holds.
* **"Set an entry target on this at the level you'd consider reasonable, and explain why"** — the agent proposes, you decide.
Because your agent sees your watchlists during its monitoring passes and when writing your [Morning Brief](/docs/automations/morning-brief), a well-kept watchlist directly improves what it surfaces for you.
## Organizing ideas [#organizing-ideas]
There's no single right structure, but these patterns work well:
* **Research queue** — things you've seen mentioned but haven't looked into yet. Ask the agent to work through it.
* **Next buys** — vetted ideas with entry targets set, waiting for price.
* **Competitors** — peers of something you hold, so you notice when the relative picture shifts.
Deeper work on an idea belongs in your [Notebook](/docs/memory/notebook) — Explore and Analyze save ideas and analyses there with stable NB-numbers, while the watchlist stays the lightweight price-and-target layer on top.
## Limits on the free plan [#limits-on-the-free-plan]
On the free plan, watchlists and [portfolios](/docs/data/portfolios) share a combined limit of 5 collections. See [Limits](/docs/reference/limits) for the full table.
## A note on targets [#a-note-on-targets]
An entry target is your line, not a recommendation. Your agent helps you research and watch levels — it does research, not financial advice, and it will never tell you what to buy or sell.
## Related [#related]
- [Monitoring & alerts](/docs/automations/monitoring-and-alerts)
- [Portfolios](/docs/data/portfolios)
- [Research](/docs/capabilities/research)
---
# Connect your portfolio
The single highest-leverage setup step. Once your holdings are connected, everything changes character: your Morning Brief talks about *your* positions, research answers weigh *your* exposure, and monitoring watches *your* portfolio for big moves — not a generic market feed.
BlockMind uses every connection for **balance-only access**. It reads balances and positions and never initiates trades, withdrawals, or transfers. Wallets share only a public address; BlockMind never asks for private keys or seed phrases. Eight supported exchanges offer provider-scoped read-only API keys. Firi does not, so BlockMind enforces balance-only behavior for Firi in the application.
## Where it all lives [#where-it-all-lives]
Manage every connection from the **Wallets & Exchanges** page — in the Pages group of the sidebar, or directly at [blockmind.app/elite/wallet](https://blockmind.app/elite/wallet). That's where you add wallets, connect exchanges, import CSVs, and manage what you've connected.
You can also just tell your agent "help me connect my wallet" in chat and it walks you through it.
## Wallets [#wallets]
Two ways to add a wallet:
1. **Connect a WalletConnect-compatible wallet** — approve the connection from your wallet app.
2. **Paste any public address** — no wallet app needed, no approval, nothing signed. If you can see the address, you can track it.
Either way, BlockMind only ever reads public on-chain data.
### Supported chains [#supported-chains]
| | |
| --------- | --------------- |
| Ethereum | Base |
| Polygon | Arbitrum |
| Optimism | BNB Smart Chain |
| Avalanche | Solana |
### Spam and dust [#spam-and-dust]
Wallets accumulate junk — airdropped spam tokens and dust balances. BlockMind filters suspicious tokens out by default so they don't pollute your portfolio value. There's a toggle if you want to see everything.
### Re-scan [#re-scan]
Holdings changed and the numbers look stale? Re-scan the wallet from the Wallets & Exchanges page to refresh its balances.
## Exchanges [#exchanges]
Nine exchanges connect via API keys:
| | | |
| ---------- | -------- | ------ |
| Binance | Coinbase | Kraken |
| OKX | KuCoin | MEXC |
| Crypto.com | Bitstamp | Firi |
You create an API key in your exchange account, paste it into BlockMind, and your balances appear.
For Binance, Coinbase, Kraken, OKX, KuCoin, MEXC, Crypto.com, and Bitstamp, grant the API key **read-only permission only**. Never enable trading or withdrawals. Firi does not offer a provider-scoped read-only key; BlockMind only reads balances and never initiates trades, withdrawals, or transfers, but the restriction is enforced by BlockMind rather than by Firi's key scope. All exchange credentials are stored encrypted.
## CSV import [#csv-import]
Holdings somewhere that doesn't connect directly? Export a CSV and import it from the Wallets & Exchanges page. Useful for smaller exchanges, cold storage records, or a spreadsheet you already maintain.
## Manual portfolios [#manual-portfolios]
You can also build a portfolio by hand — add assets and amounts yourself. Good for tracking a hypothetical portfolio, a friend's-eye view of a strategy, or positions you'd rather not link to any account.
## After connecting [#after-connecting]
* Ask your agent for a first read: "Walk me through my portfolio — what stands out?"
* Your Morning Brief and background monitoring pick up the new holdings automatically.
* The [Portfolios page](/docs/data/portfolios) gives you the multi-portfolio overview with drill-down.
## Something not working? [#something-not-working]
Balances missing, a chain not showing, an exchange key rejected — see [Troubleshooting](/docs/reference/troubleshooting), or email [contact@blockmind.app](mailto:contact@blockmind.app).
## Next steps [#next-steps]
- [Your first conversation](/docs/getting-started/first-conversation)
- [Portfolios](/docs/data/portfolios)
- [Portfolio analysis](/docs/capabilities/portfolio-analysis)
---
# Create your agent
Creating your agent takes a few minutes end to end. Here's exactly what happens, so nothing at checkout or during the boot surprises you.
## Before you start [#before-you-start]
You don't need anything prepared. Wallet addresses and exchange keys can be added later — the agent is useful from the first conversation, and gets better once it can see your portfolio.
## The steps [#the-steps]
### 1. Name your analyst [#1-name-your-analyst]
Go to [blockmind.app](https://blockmind.app) and start. The first question is the only one:
> **Give your analyst a name.**
Up to 30 characters. Choose with a little care — the name is permanent, and it's what your agent goes by everywhere: chat, briefs, emails, Telegram.
### 2. Verify your email [#2-verify-your-email]
Enter your email and type in the 6-digit code we send, or continue with Google. No passwords — BlockMind sign-in is always a code or Google.
### 3. Start Pro [#3-start-pro]
Checkout opens next. The terms in short (full detail on [Plans & pricing](/docs/overview/plans-and-pricing)):
* A card is required; you're charged **$1 today** for your first 7 days.
* After 7 days, Pro continues at the option you choose: **$29/month** or **$279/year**, unless you cancel.
* One $1 first week per account, ever.
Stripe shows your exact amount and renewal date before you confirm.
### 4. Watch it boot [#4-watch-it-boot]
Your agent's private workspace is created — it gets its own computer. The boot takes about two minutes, and the screen walks you through what's happening while its 13 starting capabilities are installed.
### 5. Say hello [#5-say-hello]
When the boot finishes, your agent sends the first message. It will introduce itself and help you get set up. From here, see [Your first conversation](/docs/getting-started/first-conversation).
Your **Morning Brief is enabled automatically** and is configured for around 8:00 in your timezone. See [Morning Brief](/docs/automations/morning-brief).
## Registering without an agent [#registering-without-an-agent]
You can also register for free without creating an agent: the classic platform includes portfolio tracking, watchlists, and market indicators at no cost. See [the classic platform](/docs/classic/overview). You can start Pro later from [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade).
## If something goes wrong [#if-something-goes-wrong]
* **Checkout finished but no agent?** The page waits for your access to activate and then forwards you automatically. If it seems stuck, retry from [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade).
* **No verification code?** Check spam, wait a minute, request a new code.
More in [Troubleshooting](/docs/reference/troubleshooting).
## Next steps [#next-steps]
- [Your first conversation](/docs/getting-started/first-conversation)
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
- [Tour the dashboard](/docs/getting-started/tour)
---
# Your first conversation
You never start from a blank box. When your agent finishes booting, **it sends the first message** — introducing itself and suggesting where to begin. Your only job is to answer.
This page covers what to say back, and the handful of chat mechanics worth knowing on day one.
## If you're not sure what to ask [#if-youre-not-sure-what-to-ask]
Two places give you ready-made starting points:
* **The Quick start rail on Home** — a row of prompts right under the composer. Click one and it fills the chat for you.
* **The Use cases catalog** — 36 prompts across 6 categories, from portfolio questions to on-chain digging. Browse it from the sidebar, or read more in [Use cases](/docs/capabilities/use-cases).
Both are just prefilled messages. Edit them, make them yours, and send.
## Five good first prompts [#five-good-first-prompts]
Concrete beats clever. These five each exercise a different part of what your agent does:
1. **Read your portfolio** — "Give me an honest read of my portfolio: what's my biggest concentration risk right now?" (Works best after you [connect your portfolio](/docs/getting-started/connect-your-portfolio).)
2. **Research something you hold** — "Research my largest holding: what's driving the price this month, and what's the bear case?"
3. **Set an alert in plain language** — "Alert me if BTC moves more than 8% in a day." Your agent turns it into a durable rule; see [Monitoring & alerts](/docs/automations/monitoring-and-alerts).
4. **Teach it a preference** — "Remember this: I invest on a 2–5 year horizon and I don't touch leverage." It saves that to memory and factors it into future answers.
5. **Shape your briefs** — "More macro context in my Morning Briefs, and always flag anything moving in my watchlist." Your agent remembers the preference and shapes future briefs around it.
## What you'll see while it works [#what-youll-see-while-it-works]
Longer questions take real work — reading data, browsing pages, running analysis. While that happens, the chat shows **live status labels** telling you what your agent is doing right now: thinking, browsing, analyzing, and so on. You're never staring at a silent spinner wondering if anything is happening.
Answers stream in as they're written, and streams survive page reloads — if you close the tab and come back, the response picks up where it was.
## Redirecting a running answer [#redirecting-a-running-answer]
There's no stop button — a running response always completes. If it's heading somewhere you didn't intend, just send a follow-up message: your agent reads it as soon as the current answer finishes and adjusts course. "Actually, skip the price history — I only care about the tokenomics" works fine.
## "Remember this" [#remember-this]
Anything worth keeping, say so: "remember this" tells your agent to store it in its memory — preferences, context, decisions. For notes you both want to read and edit later, it uses the shared [Notebook](/docs/memory/notebook). It asks before saving; nothing lands in the Notebook without you knowing.
## Attaching files [#attaching-files]
Drop files straight into the composer — up to 5 per message, 4 MB each. Supported: PDF, DOCX, XLSX, CSV, JSON, Markdown, plain text, and images (PNG, JPEG, GIF, WebP). You can also paste screenshots directly.
Good first uses: a screenshot of a chart you want a second opinion on, or a CSV of positions you'd like it to look at.
## Asking about your portfolio [#asking-about-your-portfolio]
Once your wallets or exchanges are connected, every answer is grounded in what you actually hold — "why is my portfolio down today?" gets a real answer about *your* positions, not a generic market summary. Connecting takes a few minutes and BlockMind uses the connection only to read balances and positions: [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
Your agent does research, not financial advice. It will never tell you what to buy or sell, and it can never touch your funds — the decision is always yours.
## Next steps [#next-steps]
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
- [Tour the Elite dashboard](/docs/getting-started/tour)
- [Browse all 36 use cases](/docs/capabilities/use-cases)
---
# Tour the Elite dashboard
Your Elite home lives at [blockmind.app/elite](https://blockmind.app/elite) — it's where you land after sign-in. This page walks through everything on it, top to bottom, so nothing feels mysterious on day one.
## The hero launcher [#the-hero-launcher]
At the top: a time-of-day greeting and the launcher — the fastest way into a conversation.
* If you have a recent session, you see a **preview of your latest conversation** and can jump straight back in.
* Otherwise you get a fresh **"Ask your analyst"** composer — type and go.
* **History** opens your past sessions; **New chat** starts a clean one. Sessions are date-grouped and searchable, and you can rename them.
## Quick start rail [#quick-start-rail]
Right under the launcher: a rail of ready-made prompts. Click one to prefill the composer — edit it or send as-is. It's the cure for "what do I even ask?"; the fuller catalog is [Use cases](/docs/capabilities/use-cases), also in the sidebar.
## Widgets [#widgets]
Below the launcher, your dashboard widgets. The **Morning Brief widget** is the anchor — your daily brief lands here first each morning.
Beyond that, the dashboard holds **up to 11 widgets**, and you choose them: open **Customize Dashboard** ([blockmind.app/elite/widgets](https://blockmind.app/elite/widgets)) to add, remove, and arrange. Options include the Morning Brief, Agent Feed, a live chart, portfolio and watchlist summaries, your Notebook, Capabilities, and more.
You can also just ask: "add a chart widget for ETH to my dashboard" — rearranging the dashboard is one of your agent's capabilities.
## The lane rail: your research journey [#the-lane-rail-your-research-journey]
Four lanes run the structured research flow — **Explore → Analyze → Verdicts → Track**:
| Lane | What it's for |
| -------- | -------------------------------------------------------------- |
| Explore | Guided market scans to surface ideas |
| Analyze | One-button structured analysis of a saved idea |
| Verdicts | Second opinions on an analysis from distinct expert frameworks |
| Track | Ongoing snapshots of your positions and tracked assets |
How the lanes fit together is its own page: [The research journey](/docs/capabilities/research-journey). And throughout: it's research, never advice — the agent won't tell you what to buy or sell.
## The sidebar [#the-sidebar]
Below Home, the four journey lanes, and Use cases, the sidebar has a **Pages** group:
* **Portfolios** — multi-portfolio overview with drill-down.
* **Watchlists** — assets you're tracking, with entry targets and target alerts.
* **My Agent** — your agent's capabilities: what's installed, what needs setup, and Create Capability for custom workflows.
* **Wallets & Exchanges** — every portfolio connection in one place; see [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
### The Usage pill [#the-usage-pill]
The sidebar also shows your **usage** as a percentage remaining. It resets monthly, and Morning Briefs don't draw from it. Details in [Usage](/docs/billing/usage).
## Theme [#theme]
Elite comes in teal-accented **dark and light** themes — there's a toggle to switch. Pick whichever your eyes prefer when the brief arrives.
## Pages behind direct URLs [#pages-behind-direct-urls]
Three pages aren't in the sidebar but are always a URL away:
| Page | URL |
| -------- | -------------------------------------------------------------------- |
| Settings | [blockmind.app/elite/settings](https://blockmind.app/elite/settings) |
| Activity | [blockmind.app/elite/activity](https://blockmind.app/elite/activity) |
| Notebook | [blockmind.app/elite/notebook](https://blockmind.app/elite/notebook) |
Settings covers account-level options, Activity is where in-app alerts and updates land, and the [Notebook](/docs/memory/notebook) is the shared memory you and your agent both read and write.
## Next steps [#next-steps]
- [Your first conversation](/docs/getting-started/first-conversation)
- [The research journey](/docs/capabilities/research-journey)
- [Connect your portfolio](/docs/getting-started/connect-your-portfolio)
---
# How your agent remembers
Most AI chat resets to zero every time you open it. Your agent doesn't. It carries what it learns about you forward — your preferences, your portfolio context, your open theses, the decisions you've made — so the hundredth conversation is better than the first.
Memory works in three layers.
## Layer 1: session context [#layer-1-session-context]
The conversation you're in right now. Everything said in the current chat session — your questions, the agent's research, files you've attached — is available to the agent for the rest of that session. Ask a follow-up and it knows exactly what you're both talking about.
Session context is the shortest-lived layer. When something from a conversation deserves to outlive it, it moves into one of the next two layers.
## Layer 2: your agent's own memory [#layer-2-your-agents-own-memory]
This is the agent's private, long-term memory. It has two parts:
* **Working notes** — the agent's day-to-day record of what it did and learned while working for you.
* **Curated long-term memory** — a distilled picture of you: your preferences ("short, numbers-first answers"), your portfolio context ("core position in ETH, experimenting with L2s"), and the decisions you've made along the way.
You shape this memory directly in conversation:
* **Say "remember this" and it sticks.** "Remember that I never invest in tokens under $50M market cap" becomes part of how the agent thinks about every future request.
* **Corrections stick too.** "Actually, I sold that in May" — the agent updates its picture of your holdings and stops treating the old fact as true.
Two things happen behind the scenes:
* **A weekly curation pass consolidates automatically.** Once a week, the agent reviews its working notes and folds what matters into long-term memory, so important facts don't get buried under day-to-day noise.
* **It can search its own past semantically.** Ask "what did we conclude about that gaming token a while back?" and the agent can find the earlier work by meaning, not just exact words.
## Layer 3: the shared Notebook [#layer-3-the-shared-notebook]
The [Notebook](/docs/memory/notebook) is durable memory that *both of you* read and write — saved ideas, analyses, frameworks, and verdicts, each with a stable NB-number you can reference in chat. It's where a conversation's conclusions become a record you can return to.
## Memory vs. Notebook [#memory-vs-notebook]
The agent's memory and the Notebook solve different problems:
| | Your agent's memory | The Notebook |
| ------------------- | -------------------------------- | ---------------------------------------------------------------------------------------------- |
| Who sees it | Private to the agent | Both of you — read and edit |
| How it's maintained | Automatic (plus "remember this") | Curated together; the agent asks before saving |
| Where it's visible | Surfaces in the agent's answers | At [blockmind.app/elite/notebook](https://blockmind.app/elite/notebook) and a dashboard widget |
A rough rule: memory is *how the agent understands you*; the Notebook is *what you've decided together*.
## Why this matters [#why-this-matters]
* **Answers get more yours over time.** The agent stops giving generic research and starts answering the way you'd want a long-time analyst to — in your terms, aware of your positions and risk tolerance.
* **Briefs and research reflect your open theses.** Your [Morning Brief](/docs/automations/morning-brief) and research sessions pick up where your thinking left off, instead of starting from a blank page.
* **Nothing is lost between sessions.** Close the tab, come back next week — the context is still there.
Everything in these layers is private to your workspace. Your conversations, memory, and Notebook are never shared with other users.
## Related [#related]
- [The Notebook](/docs/memory/notebook)
- [Your first conversation](/docs/getting-started/first-conversation)
- [Morning Brief](/docs/automations/morning-brief)
---
# The Notebook
The Notebook is the memory you and your agent keep *together*: durable notes both of you can read and edit. Where [your agent's own memory](/docs/memory) works silently in the background, the Notebook is explicit and visible — the written record of your ideas, analyses, and decisions.
Find it at [blockmind.app/elite/notebook](https://blockmind.app/elite/notebook), or add the Notebook widget to your dashboard for a quick view from Home.
## NB-numbers [#nb-numbers]
Every note gets a stable number like **NB-128** the moment it's created. That number never changes, which makes notes easy to reference in chat — for both of you:
* "Add this to NB-128."
* "What did we decide in NB-121?"
* "Compare this against the framework in NB-97."
The agent uses NB-numbers the same way, so when it says "I've updated NB-128 with the new levels," you know exactly where to look.
## Note kinds [#note-kinds]
Notes are typed by what they capture and where they come from:
| Kind | What it holds | Where it usually comes from |
| ------------- | ------------------------------------------------------- | --------------------------------------------------------- |
| **Note** | Anything worth keeping — a thesis, a reminder, a source | Created manually or saved from chat |
| **Idea** | A candidate worth a closer look | Saved from [Explore](/docs/capabilities/research-journey) |
| **Analysis** | A structured deep look at one idea | Saved from Analyze |
| **Framework** | Your own decision process, written down | You, working with the agent |
| **Verdict** | A second-opinion judgement on an analysis | Saved from Verdicts |
Framework notes do double duty: any framework you save in the Notebook appears as an expert on the Verdicts page, so your own decision process weighs in alongside the built-in experts. More on that in [the research journey](/docs/capabilities/research-journey).
Verdicts and analyses in the Notebook are research judgements, not financial advice — your agent will never tell you what to buy or sell, and the final call is always yours.
## Threads and Open notes [#threads-and-open-notes]
For work that spans days or weeks — evaluating a sector, tracking a thesis through a market cycle — group related notes into a **Thread**. A thread keeps the whole project in one place instead of scattered across the Notebook.
**Open notes** are the ones still in play: questions you haven't answered, theses you're still testing. They're your active desk, separate from the settled record.
## Working with the Notebook [#working-with-the-notebook]
* **Search** finds notes by content, so you don't need to remember NB-numbers.
* **Archive** notes you're done with; they're out of the way but not gone — **restore** them any time.
* **Create manually**: you can add notes and threads yourself in the Notebook, not just through chat.
* **Edit freely**: both you and the agent can update any note, so the record stays current as your thinking evolves.
## Saving etiquette [#saving-etiquette]
The agent treats the Notebook as yours:
* **It asks before saving.** When a conversation produces something durable, the agent proposes saving it — nothing lands in the Notebook without your yes.
* **It never auto-saves during automated runs.** Morning Briefs, monitoring passes, and other background work read the Notebook for context but don't write to it. Anything new that's worth keeping waits for a conversation with you.
The Notebook is private to your workspace. Only you and your agent can see it — notes are never shared with other users.
## Related [#related]
- [How your agent remembers](/docs/memory)
- [The research journey](/docs/capabilities/research-journey)
---
# Plans & pricing
BlockMind has one paid plan — **Pro** — with monthly and yearly billing, plus a free tier on the classic platform. This page is the complete picture; checkout always shows your exact terms before you confirm anything.
## BlockMind Pro — $29/month or $279/year [#blockmind-pro--29month-or-279year]
Pro is the plan that includes your personal agent.
* **Monthly:** $29/month at the current early-member price (standard price: $39/month), locked in for your first year.
* **Yearly:** $279/year at the current early-member price, saving $69 compared with 12 months at the monthly price.
* **Your personal agent** and the full Elite dashboard: Morning Briefs, research, monitoring and alerts, Notebook memory, and custom capabilities. The optional Telegram companion is available when enabled for your workspace.
* **Full monthly usage** for research, briefs don't count against it. See [Usage](/docs/billing/usage).
* **Everything premium on the classic platform** too: full AI chat and much higher analysis limits.
## The $1 first week [#the-1-first-week]
* **A card is required.** You're charged **$1 today** for the first 7 days; after that, Pro continues at the billing option you choose: $29/month or $279/year.
* **Cancel any time during the first week** to prevent the first full-price renewal. The $1 activation charge is a real charge.
* **One $1 first week per account, ever.** If you've had a trial or subscription before, a new subscription is charged at the normal price from day one.
* Stripe's checkout shows the exact amount, renewal date, and what happens next before you confirm.
## The Free plan [#the-free-plan]
Registering without a subscription is free, and stays free:
* Portfolio tracking across wallets and exchanges (read-only), watchlists, and live charts on the classic platform.
* All four market indicators: Fear & Greed, BTC Dominance, Altcoin Season, and the Contrarian Leverage Index.
* Free limits: 5 AI chat messages per day, 3 analysis requests per month, and 5 collections (portfolios and watchlists combined).
**Your agent does not run on the Free plan.** The personal agent, Morning Briefs, monitoring, and the Elite dashboard require Pro. See [what's on the classic platform](/docs/classic/overview).
## Coming later [#coming-later]
A larger tier for power users is planned. Details and pricing aren't announced — when it ships, it will be documented here.
## Common questions [#common-questions]
### Do I need a card for the trial? [#do-i-need-a-card-for-the-trial]
Yes. A card starts your first week; you're charged $1 today and then $29/month or $279/year after 7 days, depending on the billing option you choose. This is shown clearly at checkout.
### How do I cancel? [#how-do-i-cancel]
From the app: your user menu → **Change Plan**, which opens the billing portal. Cancelling during the first week prevents the first full-price renewal; the $1 activation charge remains. See [Billing & account](/docs/billing).
### What happens to my agent if I cancel? [#what-happens-to-my-agent-if-i-cancel]
It goes to sleep — everything it learned is saved, and you can reactivate any time. See [Agent lifecycle](/docs/billing/agent-lifecycle).
### Is there a guarantee? [#is-there-a-guarantee]
Yes — if BlockMind isn't for you, contact us within your first 60 days at [contact@blockmind.app](mailto:contact@blockmind.app) and we'll refund you.
### Can I pay yearly? [#can-i-pay-yearly]
Yes. Choose $279/year at checkout. If monthly Pro is already active, open the [Pro upgrade page](https://blockmind.app/elite/upgrade) and choose **Switch to yearly**. Stripe shows any prorated credit before you confirm.
## Next steps [#next-steps]
- [Create your agent](/docs/getting-started/create-your-agent)
- [Billing & account](/docs/billing)
---
# What is BlockMind?
BlockMind Pro gives you a **personal AI analyst**. It is not a chatbot in a sidebar. It is a named agent with its own workspace that knows your portfolio, remembers your decisions, researches crypto and stocks on your behalf, and checks the market even when you don't. The free classic platform does not include an agent.
You name it once, and from then on it works for you: a Morning Brief around 8:00 by default, research on demand, monitoring passes throughout the day, and alerts when something you care about moves.
## What a day with your agent looks like [#what-a-day-with-your-agent-looks-like]
1. **Around 8:00 by default, your Morning Brief arrives.** A short, personal briefing: what moved overnight, what it means for *your* holdings, and what's worth a look today. It is published to your dashboard and sent by email; it is not delivered as chat or via Telegram.
2. **During the day — ask it anything.** "Why is my portfolio down?" "Research this token before I buy." "What are whales doing in ETH?" It answers with your live portfolio as context, streaming its work as it goes.
3. **In the background, it keeps watch.** Every few hours it scans your portfolio for big moves, your tracked assets for big daily moves and major news, and the wider market for risks. If something crosses a line you set, you get an alert in the app or by email. Telegram alerts are also available when the optional companion is enabled and connected.
4. **Over time — it gets more yours.** It remembers your preferences, your open theses, and your decisions in a shared [Notebook](/docs/memory/notebook), and you can teach it your own research frameworks and [custom workflows](/docs/capabilities/custom).
## The core ideas [#the-core-ideas]
### Your agent, your workspace [#your-agent-your-workspace]
Each agent runs in its own private workspace — its own computer, dedicated to you. Your conversations, memory, and portfolio context live there and are never shared with other users. If your subscription lapses, the agent goes to sleep with everything saved.
### Grounded in your portfolio [#grounded-in-your-portfolio]
Connect wallets and exchanges for balance-only use and every answer, brief, and alert is grounded in what you actually hold. See [Connect your portfolio](/docs/getting-started/connect-your-portfolio) for exchange-key scope details.
### Research, not tips [#research-not-tips]
Your agent does research — structured analysis, second opinions, on-chain data, market context. It will never tell you what to buy or sell, and it can never touch your funds. The final decision is always yours. More in [Trust & security](/docs/security).
## The words we use [#the-words-we-use]
A quick map of the product's vocabulary, used throughout these docs:
| Term | Meaning |
| ----------------------------- | --------------------------------------------------------------------------- |
| **Your agent / your analyst** | The AI that works for you. You choose its name. |
| **Elite** | The agent's home — the dashboard at blockmind.app/elite. Included with Pro. |
| **Pro** | The subscription plan ($29/month or $279/year) that includes your agent. |
| **Capabilities** | The things your agent can do, managed on the My Agent page. |
| **Morning Brief** | Your scheduled daily briefing, set around 8:00 by default. |
| **Notebook** | The shared memory you both read and write, with numbered notes (NB-128). |
| **Usage** | Your monthly allowance, shown as a percentage ("72% left"). |
| **Classic platform** | The free tier: portfolio tracking, watchlists, market indicators. |
## Who BlockMind is for [#who-blockmind-is-for]
* **Active crypto investors** who want research depth without spending hours on it.
* **Cross-asset investors** — the agent covers stocks, ETFs, indexes, and commodities data too.
* **Busy people with positions** who mostly want the Morning Brief and an eye kept on their portfolio.
If you just want free portfolio tracking and market indicators without an agent, the [classic platform](/docs/classic/overview) does that.
## Next steps [#next-steps]
- [What your agent can do](/docs/overview/what-your-agent-can-do)
- [Plans & pricing](/docs/overview/plans-and-pricing)
- [Create your agent](/docs/getting-started/create-your-agent)
---
# What your agent can do
A quick tour of everything your agent does, with links to the full guides. If you're evaluating BlockMind, this page is the honest inventory.
## Every day, automatically [#every-day-automatically]
**[Morning Brief](/docs/automations/morning-brief):** a personal daily briefing, scheduled around 8:00 your time by default: what moved, why it matters for your holdings, what to watch today. Published to your dashboard and sent by email.
**[Monitoring & alerts](/docs/automations/monitoring-and-alerts)** — your agent checks the market every few hours: big moves in your portfolio, big daily moves and major news on tracked assets, market-wide risks. Set alerts in plain language ("tell me if BTC crosses $100k") and get them in-app, by email, or on Telegram.
## Research on demand [#research-on-demand]
**[The research journey](/docs/capabilities/research-journey)** — the core loop: **Explore** scans markets for ideas, **Analyze** turns a saved idea into a structured report, **Verdicts** gets second opinions from four expert frameworks (or your own), and **Track** monitors what you're in.
**[Coin, narrative & equities research](/docs/capabilities/research)** — structured research on any token or protocol, on sectors and capital flows, and on public stocks with quotes and fundamentals. Cross-asset: crypto, stocks, ETFs, indexes, commodities data.
**[On-chain intelligence](/docs/capabilities/on-chain)** — whale and smart-money flows, holder analysis, and wallet profiling, plus TVL, yields, and protocol revenue.
**[KOL analysis](/docs/capabilities/kol-analysis)** — market analysis through a specific analyst's framework, up to a full multi-expert discussion with cross-critique.
**[Browser](/docs/capabilities/browser)** — your agent can visit live websites in its own browser: read dashboards, extract data, capture screenshots and page PDFs.
## Your portfolio, understood [#your-portfolio-understood]
**[Portfolio analysis](/docs/capabilities/portfolio-analysis)** — holdings, exposure, performance, and position changes over your connected portfolios. Everything the agent says is grounded in what you actually hold.
**[Portfolios](/docs/data/portfolios) & [watchlists](/docs/data/watchlists)** — multi-portfolio tracking and target-based watchlists your agent reads *and* keeps updated for you.
## It remembers [#it-remembers]
**[Memory](/docs/memory)** — your agent keeps its own working memory of your preferences, context, and decisions ("remember this" works), and curates it weekly.
**[Notebook](/docs/memory/notebook)** — shared, durable notes you both read and write, each with a stable number you can reference in chat: ideas, analyses, frameworks, verdicts.
## Wherever you are [#wherever-you-are]
**[Web chat](/docs/channels/web-chat)** — multi-session chat with attachments, inline charts, artifacts, and responses that survive page reloads.
**[Telegram](/docs/channels/telegram):** an optional companion, when enabled for your workspace, for chat and alerts on your phone: text, voice notes, photos, and files.
**[Email](/docs/channels/email)** — your agent has its own inbox: it emails your brief, can send and reply to email on request, and replying to it continues the conversation.
## Make it yours [#make-it-yours]
**[Custom capabilities](/docs/capabilities/custom)** — teach your agent a repeatable workflow in conversation; it saves it (with your approval) and reuses it.
**[Charts & dashboard](/docs/capabilities/charts-and-dashboard)** — interactive charts in chat, and a home dashboard of widgets your agent can rearrange for you.
## What it won't do [#what-it-wont-do]
By design, and worth knowing up front:
* It never tells you what to buy or sell — research, not financial advice.
* It can never trade, withdraw, or move funds. BlockMind uses connections only to read balances and positions.
* It doesn't hold custody of anything.
More in [Trust & security](/docs/security).
## Try it [#try-it]
- [Plans & pricing](/docs/overview/plans-and-pricing)
- [Create your agent](/docs/getting-started/create-your-agent)
---
# Limits at a glance
One page for every hard number in the product. Each row links to the page that explains it in context.
## Chat & attachments [#chat--attachments]
| Limit | Value | Details |
| ----------------- | ----------------------------------------------------------------------------------- | ----------------------------------- |
| Files per message | 5 | [Web chat](/docs/channels/web-chat) |
| Size per file | 4 MB | [Web chat](/docs/channels/web-chat) |
| Supported types | PDF, DOCX, XLSX, CSV, JSON, Markdown, plain text, and images (PNG, JPEG, GIF, WebP) | [Web chat](/docs/channels/web-chat) |
You can also paste screenshots directly into the composer.
## Free plan [#free-plan]
The Free plan is the classic platform — your agent doesn't run on Free.
| Limit | Value | Details |
| ---------------------------------------------- | -------------------------------------------------------------------- | ---------------------------------------------------- |
| AI chat messages | 5 per day | [Classic platform](/docs/classic/overview) |
| Analysis requests | 3 per month | [DeepDive reports](/docs/classic/deepdive-reports) |
| Collections (portfolios + watchlists combined) | 5 | [Classic platform](/docs/classic/overview) |
| Free market indicators | 4 (Fear & Greed, BTC Dominance, Altcoin Season, Contrarian Leverage) | [Market indicators](/docs/classic/market-indicators) |
## Trial & billing [#trial--billing]
| Limit | Value | Details |
| -------------------------- | ------------------------------------------------------------------------------------------------ | --------------------------------------------------- |
| First-week offer | 7 days for $1 (card required), then $29/month or $279/year based on your choice unless cancelled | [Plans & pricing](/docs/overview/plans-and-pricing) |
| $1 first weeks per account | One, ever; returning subscribers are charged the normal price from day one | [Plans & pricing](/docs/overview/plans-and-pricing) |
## Alerts & monitoring [#alerts--monitoring]
| Limit | Value | Details |
| ---------------------------- | ---------------------------------------------------------------------- | -------------------------------------------------------------- |
| Monitoring passes | Every few hours, 06:00–23:00 your time | [Monitoring & alerts](/docs/automations/monitoring-and-alerts) |
| Portfolio move threshold | >10% in 24 hours | [Monitoring & alerts](/docs/automations/monitoring-and-alerts) |
| Tracked-asset move threshold | >5% daily move (plus major news) | [Monitoring & alerts](/docs/automations/monitoring-and-alerts) |
| Alert latency | Up to a few hours — evaluated during monitoring passes, not tick-level | [Monitoring & alerts](/docs/automations/monitoring-and-alerts) |
Alerts and monitoring are research, not financial advice — your agent flags what moved, but it will never tell you what to buy or sell.
## Telegram [#telegram]
| Limit | Value | Details |
| --------------------- | ---------------- | ----------------------------------- |
| Connect link validity | 60 minutes | [Telegram](/docs/channels/telegram) |
| Chats per account | One private chat | [Telegram](/docs/channels/telegram) |
## DeepDive reports [#deepdive-reports]
| Limit | Value | Details |
| -------------------- | ----------------------------- | -------------------------------------------------- |
| Public DeepDive | 5 per day per email address | [DeepDive reports](/docs/classic/deepdive-reports) |
| Signed-in, Free plan | 3 analysis requests per month | [DeepDive reports](/docs/classic/deepdive-reports) |
| Signed-in, paid | 30 per rolling 30 days | [DeepDive reports](/docs/classic/deepdive-reports) |
## Connections [#connections]
| Limit | Value | Details |
| ------------------- | ---------------------------------------------------------------------------------- | ---------------------------------------------------------------------- |
| Supported networks | 8: Ethereum, Base, Polygon, Arbitrum, Optimism, BNB Smart Chain, Avalanche, Solana | [Connect your portfolio](/docs/getting-started/connect-your-portfolio) |
| Supported exchanges | 9 — Binance, Coinbase, Kraken, OKX, KuCoin, MEXC, Crypto.com, Bitstamp, Firi | [Connect your portfolio](/docs/getting-started/connect-your-portfolio) |
BlockMind uses connections only to read balances and positions. Eight exchanges offer provider-scoped read-only keys; Firi's restriction is enforced in the application. CSV import and manual portfolios are also available.
## Dashboard [#dashboard]
| Limit | Value | Details |
| -------------------- | -------- | ------------------------------------------------------------- |
| Customizable widgets | Up to 11 | [Charts & dashboard](/docs/capabilities/charts-and-dashboard) |
## Agent name [#agent-name]
| Limit | Value | Details |
| ----------- | ------------------------------------ | ------------------------------------------------------------ |
| Name length | 30 characters | [Create your agent](/docs/getting-started/create-your-agent) |
| Renaming | Not possible — the name is permanent | [Create your agent](/docs/getting-started/create-your-agent) |
## Related [#related]
- [Plans & pricing](/docs/overview/plans-and-pricing)
- [Usage](/docs/billing/usage)
- [Troubleshooting](/docs/reference/troubleshooting)
---
# Troubleshooting
Most issues here resolve in under a minute. If yours doesn't, email us at the bottom of this page.
## My agent is waking up or won't respond [#my-agent-is-waking-up-or-wont-respond]
Your agent sleeps between tasks. When you open the dashboard or send a message, it wakes automatically — you'll see a "Waking…" screen with your agent's name while it boots. This usually takes a moment; give it time to finish rather than reloading repeatedly.
If the wake seems stuck, use the **Retry wake** action on the same screen. That restarts the wake from scratch and resolves most hangs.
## "Your agent is paused" [#your-agent-is-paused]
This means your plan has lapsed — a payment failed or the subscription was cancelled. Nothing is lost: your agent's memory, Notebook, conversations, and portfolio context are all saved while it sleeps.
To bring it back, reactivate Pro at [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade). If you already upgraded and still see the paused screen, use **"I already upgraded — retry"** to re-check your subscription.
More on what pausing means in [Agent lifecycle](/docs/billing/agent-lifecycle).
## Restarting your agent [#restarting-your-agent]
If your agent is behaving oddly, you can restart it yourself: open [Settings](https://blockmind.app/elite/settings), scroll to the **Danger Zone**, and use **Restart** (or **Wake** if it's asleep). A restart is safe — memory and Notebook survive it.
**Delete workspace** in the same section is permanent. It removes your agent and everything it has learned, and there is no undo. See [Agent lifecycle](/docs/billing/agent-lifecycle) before using it.
## A response is stuck [#a-response-is-stuck]
Responses stream live, and streams survive page reloads and reconnects — if you see "Reconnecting…", the work is still running on your agent's side. There's no stop button: a running response always completes. If it's heading the wrong way, let it finish, then send a follow-up to redirect it.
Long tasks (deep research, browser sessions, Discussion mode) genuinely take time — the per-tool status labels tell you what it's doing right now.
## My Morning Brief didn't arrive [#my-morning-brief-didnt-arrive]
Check in this order:
1. **The Home widget first.** The brief lands on your [Elite home](https://blockmind.app/elite) before anywhere else.
2. **Was your agent paused overnight?** A paused agent doesn't produce briefs. Reactivate and the next morning's brief resumes.
3. **Email can lag the widget.** Sending is queued, so the email version may arrive after the widget version. Check your spam folder too.
More in [Morning Brief](/docs/automations/morning-brief).
## Telegram problems [#telegram-problems]
* **Connect link doesn't work** — links expire after 60 minutes. Generate a fresh one from My Agent or [Settings](https://blockmind.app/elite/settings).
* **Disconnected unexpectedly** — sending /stop in Telegram disconnects the companion. Use **Reconnect** from Settings to link again.
* **Can't connect a second chat** — one private chat per account is the limit.
* **No connect card at all** — Telegram is still rolling out; if you don't see the option yet, it isn't available on your workspace yet.
More in [Telegram](/docs/channels/telegram).
## Checkout finished but I don't see my agent [#checkout-finished-but-i-dont-see-my-agent]
After checkout, the page waits for your access to activate and then forwards you to your agent. If you closed the tab or the forward didn't happen, go to [blockmind.app/elite/upgrade](https://blockmind.app/elite/upgrade) and retry from there.
## I'm out of usage [#im-out-of-usage]
The [Usage page](https://blockmind.app/elite/credits) shows how much of your monthly allowance is left and when it resets. Usage resets monthly on paid plans, and Morning Briefs never draw from it. Details in [Usage](/docs/billing/usage).
## Classic platform issues [#classic-platform-issues]
* **Wallet won't connect** — refresh the page, make sure your wallet extension is unlocked, and try another browser if it persists.
* **Tokens missing from a wallet** — re-scan the wallet to refresh balances, and check the suspicious-tokens toggle: spam and dust are filtered by default, and a real token can occasionally be caught.
* **Sign-in code never arrives** — check your spam folder, wait a minute or two, then request a new code.
## Still stuck? [#still-stuck]
Email **[contact@blockmind.app](mailto:contact@blockmind.app)** with what you tried and, if possible, your agent's name and roughly when the problem happened. We read everything.
## Related [#related]
- [Limits at a glance](/docs/reference/limits)
- [Agent lifecycle](/docs/billing/agent-lifecycle)
- [Usage](/docs/billing/usage)
---
# Trust & security
BlockMind is built around a strict product boundary: **the application reads balances and positions and never initiates trades, withdrawals, or transfers.** Your agent has no execution tools. This page explains that model, the important Firi credential exception, and what we store.
## Balance-only by design [#balance-only-by-design]
When you connect a wallet or an exchange, or paste a public address:
* Wallet connections share a public address. We never receive private keys or seed phrases and cannot sign wallet transactions or approve contracts.
* For Binance, Coinbase, Kraken, OKX, KuCoin, MEXC, Crypto.com, and Bitstamp, use a **provider-scoped read-only API key** with trading and withdrawals disabled.
* Firi does not offer a read-only API-key scope. BlockMind stores the credentials encrypted and enforces balance-only behavior in the application; it never initiates trades, withdrawals, or transfers.
Provider-scoped read-only keys and public wallet addresses cannot authorize fund movement. Firi credentials deserve extra care because Firi does not enforce that restriction at key level. Revoke any exchange credential immediately if you suspect your exchange account or BlockMind account has been compromised.
Wallet connections use the industry-standard WalletConnect protocol and share only your public address. Hardware wallets work the same way. You can remove a connected wallet or exchange at any time.
## Your agent's boundaries [#your-agents-boundaries]
Your agent is an analyst, not a trader:
* **It does research, not financial advice.** It will not tell you what to buy or sell. Its analyses and verdicts are inputs to *your* decision.
* **It cannot execute trades** — even if you ask it to. There is no execution capability of any kind.
* **AI can be wrong.** Analyses may be inaccurate or outdated; always verify before acting on anything important.
## Your workspace is private [#your-workspace-is-private]
Each agent runs in its own private workspace, dedicated to one account:
* Your conversations, your agent's memory, your Notebook, and your portfolio context live in your workspace and are **not shared with other users**.
* When your agent sleeps (or your plan lapses), the workspace's data is saved, not shared or repurposed.
## What we store [#what-we-store]
* Your email address, for sign-in and for briefs.
* Portfolio and watchlist data — the addresses you connect and the holdings we read from them.
* Your conversations with your agent, its memory, and your Notebook.
* Documents you attach in chat.
* Exchange API credentials you provide, encrypted at rest. Eight supported exchanges provide read-only key scopes; Firi's balance-only restriction is enforced by BlockMind in the application.
## Who processes your data [#who-processes-your-data]
We use a small set of service providers to run BlockMind: payment processing (Stripe), cloud hosting, AI model providers accessed through a managed gateway, web-search providers, the agent's email service, and a cloud-browser service. Providers process data only to deliver their function.
## Analytics, honestly [#analytics-honestly]
We use product analytics tied to your account to understand how BlockMind is used and improve it. Our marketing pages also use advertising pixels (such as Meta's) for ad measurement.
## Signing in [#signing-in]
* **Passwordless**: a 6-digit code to your email, or Google sign-in. There's no password to steal or reuse.
* Sessions expire and refresh automatically; signing in again is always just a code away.
## Deleting your data [#deleting-your-data]
Email [contact@blockmind.app](mailto:contact@blockmind.app) from your account email to request deletion of your account and data. You can also delete your agent's workspace yourself at any time — see [Agent lifecycle](/docs/billing/agent-lifecycle).
## Reporting a security concern [#reporting-a-security-concern]
Found a vulnerability? We appreciate responsible disclosure: email [contact@blockmind.app](mailto:contact@blockmind.app) with a description, steps to reproduce, and how to reach you.
## Common questions [#common-questions]
### Can BlockMind or my agent steal my crypto? [#can-blockmind-or-my-agent-steal-my-crypto]
BlockMind and your agent do not initiate trades, withdrawals, transfers, or wallet signatures. Use provider-scoped read-only exchange keys wherever they are available. Firi is the exception: its key is not scopeable as read-only, so BlockMind enforces balance-only behavior in the application.
### What if BlockMind gets hacked? [#what-if-blockmind-gets-hacked]
An attacker could expose the data listed under "What we store." Public wallet addresses and provider-scoped read-only exchange keys cannot authorize fund movement. Firi credentials are more sensitive because Firi does not offer a read-only scope; revoke them promptly if you suspect a compromise.
### Can I use BlockMind anonymously? [#can-i-use-blockmind-anonymously]
You need a working email (briefs and sign-in codes go there), but any email service works. Your wallet addresses and portfolio data live in your private workspace and aren't shared with other users.
### Should I use a hardware wallet? [#should-i-use-a-hardware-wallet]
For significant holdings, yes — and it changes nothing about how BlockMind works, since we only ever read public addresses.
---
## FAQ
# Frequently Asked Questions
Quick answers to the questions we hear most. For depth, every answer links to the full guide.
## The agent [#the-agent]
### What is BlockMind? [#what-is-blockmind]
BlockMind Pro gives you a personal AI analyst: a named agent with its own workspace that researches crypto, stocks, and markets for you, publishes a [Morning Brief](/docs/automations/morning-brief) to your dashboard and sends it by email around 8:00 by default, watches your portfolio between conversations, and remembers what matters to you. The free classic platform does not include an agent. See [What is BlockMind?](/docs/overview/what-is-blockmind)
### Will it tell me what to buy? [#will-it-tell-me-what-to-buy]
No. Your agent does research, not financial advice — analyses, second opinions, on-chain data, and context, so *you* can decide. That's a deliberate boundary, not a limitation we plan to remove.
### What can it actually do? [#what-can-it-actually-do]
Daily briefs, research on any token or stock, structured analysis with expert verdicts, on-chain intelligence, monitoring and price alerts, a browser for live websites, persistent memory, and custom capabilities you teach it. Where available, the Telegram companion adds mobile chat and alerts. The full inventory: [What your agent can do](/docs/overview/what-your-agent-can-do).
### Does it remember things? [#does-it-remember-things]
Yes — that's most of the point. Say "remember this" and it does. Durable work is saved in a shared [Notebook](/docs/memory/notebook) with numbered notes you can reference later ("what did we decide in NB-121?").
### Can I talk to it from my phone? [#can-i-talk-to-it-from-my-phone]
The web app works on mobile browsers. If the [Telegram companion](/docs/channels/telegram) is available on your workspace, you can also connect it for the same conversation, including voice notes, photos, files, and alerts. Telegram is optional and does not deliver Morning Briefs; those arrive on your dashboard and by email.
## Pricing & trial [#pricing--trial]
### What does BlockMind cost? [#what-does-blockmind-cost]
**Pro is $29/month or $279/year** at the current early-member price and includes your personal agent and everything on the platform. The standard monthly price is $39. The classic platform — portfolio tracking, watchlists, market indicators — is free. Full details: [Plans & pricing](/docs/overview/plans-and-pricing).
### Is there a trial? Do I need a card? [#is-there-a-trial-do-i-need-a-card]
Your first 7 days of Pro cost **$1**, and a card is required. After 7 days, Pro continues at the option you choose: $29/month or $279/year. Cancel before the first week ends and there is no further charge. One introductory week per account, ever.
### How do I cancel? [#how-do-i-cancel]
User menu → **Change Plan** opens the billing portal. See [Billing & account](/docs/billing).
### What happens to my agent if I stop paying? [#what-happens-to-my-agent-if-i-stop-paying]
It goes to sleep — everything it learned is saved. Reactivate any time and it picks up where it left off. See [Agent lifecycle](/docs/billing/agent-lifecycle).
### What's free, exactly? [#whats-free-exactly]
The classic platform: portfolio tracking across wallets and exchanges, watchlists, live charts, and all four market indicators (Fear & Greed, BTC Dominance, Altcoin Season, and Contrarian Leverage), plus limited AI chat (5 messages/day) and analyses (3/month). The agent itself requires Pro. See [the classic platform](/docs/classic/overview).
## Security & privacy [#security--privacy]
### Can BlockMind access my funds? [#can-blockmind-access-my-funds]
BlockMind uses wallet and exchange connections only to read balances and positions, and the agent has no tools to trade, withdraw, or transfer. Wallets share public addresses, never private keys or seed phrases. Eight supported exchanges offer provider-scoped read-only API keys. Firi does not, so its balance-only restriction is enforced by BlockMind in the application. See [Trust & security](/docs/security).
### Can the agent trade for me? [#can-the-agent-trade-for-me]
No. There is no execution capability of any kind — the agent can see positions but cannot trade, withdraw, or move funds, even if you ask.
### Is my data private? [#is-my-data-private]
Your conversations, memory, Notebook, and portfolio live in your agent's private workspace and are not shared with other users. We do use product analytics and (on marketing pages) ad pixels — the honest detail is in [Trust & security](/docs/security).
### How do I delete my data? [#how-do-i-delete-my-data]
Email [contact@blockmind.app](mailto:contact@blockmind.app) from your account email. You can also delete your agent's workspace yourself in Settings.
## Portfolios & connections [#portfolios--connections]
### Which wallets and chains are supported? [#which-wallets-and-chains-are-supported]
Any WalletConnect-compatible wallet, or just paste a public address. Supported networks: Ethereum, Base, Polygon, Arbitrum, Optimism, BNB Smart Chain, Avalanche, and Solana. See [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
### Can I connect exchange accounts? [#can-i-connect-exchange-accounts]
Yes — nine exchanges: Binance, Coinbase, Kraken, OKX, KuCoin, MEXC, Crypto.com, Bitstamp, and Firi. Use provider-scoped read-only API keys for the first eight. Firi does not offer that scope, so BlockMind enforces balance-only behavior in the application. CSV import and manual portfolios work too.
### Why are some tokens missing or showing $0? [#why-are-some-tokens-missing-or-showing-0]
Very new tokens may lack price data, tiny balances are dust-filtered, and suspicious tokens are hidden by default (there's a toggle). Re-scan the wallet to refresh. More in [Troubleshooting](/docs/reference/troubleshooting).
## Product [#product]
### Is there a mobile app? [#is-there-a-mobile-app]
Not a native one yet. The web app works well on mobile browsers. Where available, the optional [Telegram companion](/docs/channels/telegram) adds on-the-go chat and alerts.
### Does BlockMind track NFTs? [#does-blockmind-track-nfts]
No — BlockMind focuses on fungible tokens and other priced assets (crypto, stocks, ETFs, indexes, commodities data).
### Can I export my data or get tax reports? [#can-i-export-my-data-or-get-tax-reports]
No. For tax reporting we recommend dedicated tools; if you need a copy of your data, email [contact@blockmind.app](mailto:contact@blockmind.app).
### What is DeepDive? [#what-is-deepdive]
Free AI research reports on the classic platform — enter a coin and get a structured report. Limits apply (they're real: see [DeepDive reports](/docs/classic/deepdive-reports)). Your agent's [research](/docs/capabilities/research) goes deeper and is grounded in your portfolio.
## Support [#support]
### How do I get help? [#how-do-i-get-help]
* Email: [contact@blockmind.app](mailto:contact@blockmind.app)
* These docs — start with [Troubleshooting](/docs/reference/troubleshooting)
* X/Twitter: [@blockmind\_agent](https://twitter.com/blockmind_agent)
### How do I report a bug or request a feature? [#how-do-i-report-a-bug-or-request-a-feature]
Email [contact@blockmind.app](mailto:contact@blockmind.app) with what you were doing, what happened, and a screenshot if you have one.
---
## Blog Posts
# 5 Signs a Crypto Project Might Be a Rug Pull
Published: 2026-02-07
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: problem-solving
Primary keyword: signs of a crypto rug pull
Keywords: crypto rug pull, rug pull signs, crypto scam warning signs
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Before investing in any crypto project, check for five red flags: anonymous teams, concentrated token supply, hype-driven marketing, missing audits, and unlocked liquidity. No single sign proves fraud, but several unresolved signs are a reason to stop and investigate before taking risk.
***
A rug pull happens when developers abandon a project and run off with investors' funds. Rug pulls are one part of a much wider fraud problem: in its [2026 Crypto Crime Report](https://www.chainalysis.com/blog/crypto-scams-2026/), Chainalysis estimated that scams received at least $14 billion on-chain in 2025, with the total likely to rise as more illicit addresses are identified. Here's how to spot the warning signs before you invest.
## 1. Anonymous Team with No Track Record [#1-anonymous-team-with-no-track-record]
**The red flag:** The team uses pseudonyms, has no LinkedIn profiles, no GitHub history, and no verifiable background in crypto or finance.
**Why it matters:** Anonymous teams have no reputation to protect. If things go wrong, they disappear without consequences.
**What to check:**
* Search team members on LinkedIn, Twitter, GitHub
* Look for past projects they've worked on
* Check if they've spoken at conferences or appeared in interviews
* Verify any claimed credentials
**The exception:** Some legitimate projects (like Bitcoin) were created by anonymous founders. But these are rare. Most successful projects have known, accountable teams.
## 2. Token Distribution Heavily Favors Insiders [#2-token-distribution-heavily-favors-insiders]
**The red flag:** A small number of wallets hold 50%+ of the token supply, and these aren't locked or vesting.
**Why it matters:** If insiders can dump their tokens at any time, they can crash the price and exit with profits while retail investors are left holding worthless tokens.
**What to check:**
* Use a blockchain explorer to see top holders
* Check if team tokens are locked with vesting schedules
* Look for wallet concentration metrics
* Verify any claimed locks on-chain (not just in the whitepaper)
**Healthy range:** Team and insider allocations should ideally be under 20%, with multi-year vesting schedules and transparent lock addresses.
## 3. Unrealistic Promises and Hype-Driven Marketing [#3-unrealistic-promises-and-hype-driven-marketing]
**The red flag:** The project promises guaranteed returns, "100x potential," or claims to revolutionize everything without clear technical details.
**Why it matters:** Real projects focus on building, not hyping. Excessive marketing with little substance is a classic pump-and-dump setup.
**Warning phrases:**
* "Guaranteed returns" (nothing in crypto is guaranteed)
* "Next Bitcoin/Ethereum killer"
* "Get in before it's too late"
* "Elon Musk / celebrity is involved" (almost always fake)
**What real projects do:** They publish technical documentation, show working code, partner with known entities, and focus on solving real problems.
## 4. No Audit or Fake Audit [#4-no-audit-or-fake-audit]
**The red flag:** The smart contract hasn't been audited, or the "audit" is from an unknown firm that rubber-stamps everything.
**Why it matters:** Unaudited contracts can contain hidden functions that let developers drain funds. Some fake audits don't even review the deployed code.
**What to check:**
* Is there an audit from a reputable firm? (CertiK, Trail of Bits, OpenZeppelin, Consensys Diligence)
* Does the deployed contract match the audited code?
* Were critical issues found and fixed?
* Can you find the audit report on the auditor's website (not just the project's)?
**Red flag:** A project that claims to be audited but won't share the report, or links to a report that doesn't match the current contract.
## 5. Liquidity Can Be Removed by the Team [#5-liquidity-can-be-removed-by-the-team]
**The red flag:** The project's liquidity pool isn't locked, meaning developers can withdraw it at any time.
**Why it matters:** When liquidity disappears, you can't sell your tokens. The price crashes to zero instantly.
**What to check:**
* Is liquidity locked? (via Unicrypt, Team Finance, or similar)
* How long is the lock? (Under 6 months is suspicious)
* Who owns the liquidity tokens?
* Is there a timelock on removing liquidity?
**The setup:** Scammers create a token, add liquidity so people can buy, wait for price to rise, then remove all liquidity and disappear.
## How BlockMind Helps You Stay Safe [#how-blockmind-helps-you-stay-safe]
BlockMind is a personal AI investing agent for crypto, stocks, and commodities. With a Pro trial or subscription, your agent works in its own workspace, researches tokens on demand, and keeps watching after you buy:
* **Research on demand:** Ask your agent to investigate holder concentration, on-chain activity, team background, and fundamentals. It reports back in chat with the evidence behind its conclusions.
* **Monitoring while you sleep:** Your agent watches your holdings and brings material changes into a Morning Brief delivered to your dashboard and email.
* **Expert verdicts:** A committee of expert frameworks, including a dedicated risk lens, reviews saved analyses and returns a call: approved, watch, wait, or reject.
BlockMind uses wallet and exchange connections only to read balances and positions. The agent can't trade, withdraw, or move funds even if you told it to.
If you only want a one-off first pass, run a free [DeepDive report](https://blockmind.app/deepdive). It gives you a structured research view of the project, market data, and holder distribution. It does not replace checking audit reports on the auditor's own website or confirming liquidity locks on-chain.
Understanding market sentiment also helps you avoid hype-driven traps. When the [Fear & Greed Index](/blog/fear-greed-index-guide) shows extreme greed, weak projects can attract capital before investors examine the details. Learn [what rug-pull checkers inspect and miss](/blog/crypto-rug-pull-checker-guide), then use the broader [pre-buy crypto checklist](/blog/what-to-check-before-buying-crypto) instead of treating one score as a safety certificate.
## Frequently Asked Questions [#frequently-asked-questions]
### What is a crypto rug pull? [#what-is-a-crypto-rug-pull]
A rug pull is a type of crypto scam where developers create a token, attract investors, then abandon the project and take investors' funds. It's called a "rug pull" because the floor (liquidity) is literally pulled from under investors.
### How common are rug pulls? [#how-common-are-rug-pulls]
Rug pulls remain a recurring risk, particularly among new tokens on decentralized exchanges, but reliable global totals vary because researchers classify scams differently. For broader context, Chainalysis found at least $14 billion in on-chain scam inflows in 2025.
### Can I get my money back after a rug pull? [#can-i-get-my-money-back-after-a-rug-pull]
In most cases, no. Because rug pulls happen on decentralized, permissionless platforms, there's usually no central authority to recover funds. This is why prevention and due diligence are critical.
### How can I check if a token is a potential rug pull? [#how-can-i-check-if-a-token-is-a-potential-rug-pull]
Use BlockMind's free DeepDive report for a first pass on the project, market data, and holder distribution, or ask your Pro agent to investigate further. DeepDive does not verify audit status or liquidity locks. Always find the audit report on the auditor's own website, inspect holders on a blockchain explorer, and confirm liquidity locks on-chain yourself.
## The Bottom Line [#the-bottom-line]
No single red flag means a project is definitely a scam. Multiple warning signs should make you very cautious. The best protection is simple:
1. **Research before investing:** A short check can prevent a costly mistake.
2. **Start small:** Never invest more than you can afford to lose.
3. **Verify claims independently:** Don't trust, verify.
4. **Walk away from impossible promises:** If it seems too good to be true, it probably is.
The crypto space has legitimate projects, but it also has predators targeting uninformed investors. Stay informed, stay skeptical, and protect your capital.
## Sources [#sources]
1. [Chainalysis: 2026 Crypto Scam Trends](https://www.chainalysis.com/blog/crypto-scams-2026/), January 2026.
2. [SEC Investor.gov: Crypto Asset Scam Warning Signs](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/crypto-scams), May 2024.
3. [CFTC: Customer Advisory on Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html), accessed July 2026.
---
# AI Crypto Agent vs Trading Bot: Research and Execution Are Not the Same
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: comparison
Primary keyword: AI crypto agent vs trading bot
Keywords: AI crypto agent vs trading bot, crypto trading bot vs AI agent, AI crypto agent, automated crypto trading, crypto research agent
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** An AI crypto agent gathers and interprets information; a trading bot sends orders according to rules or model output. Some products combine both, but the decisive difference is authority—not whether they use AI. Before connecting anything, identify what it can read, what it can decide, and what it can execute.
***
The phrase “AI crypto agent” is often used for everything from a chatbot to fully automated trading software. That makes comparisons confusing and can obscure the most important fact: **can the system move money?**
For the wider product-selection landscape, start with our methodology-led comparison of the [best AI crypto research tools](/blog/best-ai-crypto-research-tools). This article owns the narrower boundary between research authority and execution authority.
A research agent may read public data, analyze a portfolio, save a thesis, and alert you. A trading bot normally has API permission to place or cancel orders. An execution agent may go further by deciding and acting through a wallet or exchange. Those are different products with different failure modes, even if all three have a chat box.
BlockMind is on the research side of that line. Your agent can analyze connected balances and positions, but it **cannot trade, withdraw, transfer, or sign transactions**. It provides research and monitoring, never buy or sell instructions. That boundary is documented in [Trust and security](/docs/security).
## Methodology [#methodology]
This comparison classifies product categories by their documented authority: data they can read, outputs they produce, credentials they require, and actions they can execute. It uses regulator guidance and official exchange permission documentation current to July 10, 2026. It is not a performance test, backtest, or hands-on product review, and it does not rank returns. Product labels vary, so the permission screen and legal terms take precedence over the words “agent” or “bot.”
## AI crypto agent vs trading bot at a glance [#ai-crypto-agent-vs-trading-bot-at-a-glance]
| Question | Research-oriented AI crypto agent | Rules-based trading bot | Execution-capable AI agent |
| ------------------------- | ----------------------------------------------------------------- | --------------------------------------------------- | -------------------------------------------------- |
| Primary job | Find, synthesize, explain, monitor | Execute predefined trading logic | Interpret a goal, decide actions, and execute |
| Typical input | Questions, portfolio context, documents, market and on-chain data | Indicators, price feeds, parameters, exchange state | Natural-language goal plus tools and account state |
| Typical output | Report, alert, dashboard, hypothesis | Orders, cancellations, logs | Research, decisions, orders, transactions |
| Minimum permission | Public data or read-only account access | Trade permission | Trade permission and possibly wallet signing |
| Main failure | Wrong or unsupported conclusion | Bad rule, code defect, stale data, or bad execution | Wrong reasoning plus execution failure |
| Human role | Verify evidence and decide | Design, test, supervise, and stop | Define constraints and closely supervise |
| Can it guarantee returns? | No | No | No |
The [CFTC warns](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html) that AI cannot predict the future or sudden market changes and that guaranteed or implausibly high returns are fraud red flags. FINRA similarly warns that automated trades can create losses before a user realizes what happened, particularly when an unregistered third party controls instructions sent to an account.
## The RACE model: classify authority before features [#the-race-model-classify-authority-before-features]
Use **RACE** to classify any crypto automation product:
1. **Research:** What information can it retrieve and analyze?
2. **Advice:** Does it merely describe evidence, or does it recommend a specific action?
3. **Control:** What credentials, approvals, and policy settings can it access?
4. **Execution:** Can it place orders, transfer assets, approve contracts, or sign transactions?
Marketing usually emphasizes Research and Advice. Security depends on Control and Execution.
### Level 0: public research [#level-0-public-research]
The system reads public websites, market data, and blockchains. It receives no private account data. A wrong answer can still cause a bad decision, but the system cannot act on the account.
### Level 1: portfolio-aware, read-only research [#level-1-portfolio-aware-read-only-research]
The system sees balances, positions, and perhaps transaction history through public wallet addresses or exchange permissions. Privacy exposure increases because holdings and behavior may be sensitive. Direct financial authority should remain absent.
### Level 2: trading authority [#level-2-trading-authority]
The system or bot can place and cancel orders. For example, Coinbase's [Advanced Trade API](https://www.coinbase.com/developer-platform/products/advanced-trade-api) supports automated market, limit, and stop-limit orders. Exchange permission models distinguish view access from trade access; a bot needs the latter.
### Level 3: transfer or signing authority [#level-3-transfer-or-signing-authority]
The system can withdraw, transfer, approve contracts, or sign wallet transactions. This is the highest-consequence boundary. A bug, compromised credential, malicious integration, or incorrect interpretation can directly move assets.
Do not infer the level from the word “agent,” “copilot,” or “bot.” Read the permission screen and provider documentation.
## What is an AI crypto agent? [#what-is-an-ai-crypto-agent]
An AI crypto agent is software that uses models and tools to pursue a research or operational objective across more than one step. It may plan an investigation, query market and on-chain sources, use a connected portfolio, summarize findings, remember context, and monitor conditions over time.
The useful distinction from a simple chatbot is workflow. A chatbot answers a prompt. An agent may decide which sources to query, preserve state, and continue work on a schedule. Our guide to [what an AI crypto agent is](/blog/what-is-an-ai-crypto-agent) explains that architecture at a user level.
“Agent” does not imply autonomy over funds. BlockMind's agent, for example, can:
* research tokens, protocols, narratives, and public companies;
* analyze connected portfolios for balance-only use;
* save durable context in a [Notebook](/docs/memory/notebook);
* prepare a portfolio-aware [Morning Brief](/blog/crypto-morning-brief);
* run periodic [monitoring and alerts](/docs/automations/monitoring-and-alerts).
It has no execution tool and does not tell users what to buy or sell.
## What is a crypto trading bot? [#what-is-a-crypto-trading-bot]
A crypto trading bot is software that submits orders automatically according to programmed logic. The logic may be simple—buy when one moving average crosses another—or complex, involving portfolio constraints, statistical models, or AI-generated signals.
The “bot” label describes automation, not intelligence. A deterministic script can be a bot. An AI model that selects a strategy can sit upstream of a bot. Either way, the exchange sees authenticated API calls with permissions.
Official exchange documentation makes the authority split concrete:
* [Coinbase Exchange](https://help.coinbase.com/en/exchange/managing-my-account/how-to-create-an-api-key) separates **View**, **Trade**, **Transfer**, and **Manage** permissions.
* [OKX](https://my.okx.com/docs-v5/) separates **Read**, **Trade**, and **Withdraw**.
* [Kraken](https://support.kraken.com/articles/360000919966-how-to-create-an-api-key) exposes distinct permissions for querying funds, modifying trades, and withdrawing funds.
* [Binance](https://developers.binance.com/legacy-docs/binance-spot-api-docs/rest-api/request-security) says keys can be limited to endpoint types and cannot trade by default until trading is enabled.
A portfolio tracker should not need trade or withdrawal authority. A trading bot normally needs trade authority but should not need withdrawal authority. An unexplained request for more is a reason to stop.
## Where the products overlap [#where-the-products-overlap]
Research agents and trading bots can use the same ingredients:
* current price and order-book data;
* technical indicators;
* news and social sentiment;
* derivatives data;
* portfolio positions;
* on-chain transfers;
* risk limits and alerts.
The overlap is analytical. The divergence happens at the last mile. A research agent says, “Funding turned sharply positive while open interest rose; verify whether leverage is crowded.” A bot turns a condition into an order. An execution agent may generate both the interpretation and the order.
That last configuration combines two error sources. The model can misread the situation, and the execution system can mishandle the action. Human confirmation between them is not pointless friction; it is a control.
## Different failure modes [#different-failure-modes]
### Research-agent failures [#research-agent-failures]
* Citing a source that does not support the claim
* Confusing two tokens with the same symbol
* Treating a wallet transfer as proof of intent
* Using stale tokenomics or governance information
* Overstating causation behind a price move
* Forgetting an important user constraint
The defense is traceability: contract addresses, dates, source links, uncertainty, and independent checks.
### Trading-bot failures [#trading-bot-failures]
* Overfitting a backtest to historical noise
* Ignoring fees, spread, slippage, taxes, or funding
* Using a stale or manipulated price feed
* Repeating orders after a retry or network fault
* Trading through a market-regime change
* Exceeding position or loss limits
* Leaking a trade-enabled API credential
The defense includes paper trading, small limits, idempotent order handling, monitoring, kill controls, and a permission model that excludes withdrawals.
### Execution-agent failures [#execution-agent-failures]
An execution-capable AI system inherits both lists and adds ambiguity. Natural-language goals can be underspecified. A request such as “reduce my risk” does not define which assets, maximum slippage, tax constraints, time horizon, or acceptable transaction paths.
FINRA's [2025 auto-trading warning](https://www.finra.org/investors/insights/auto-trading-unregistered-entities) highlights unsupported profitability claims, AI washing, unsuitable trades, and credential risk. Registration rules vary by product and jurisdiction, but a technology label never removes the need to verify who operates the service and what oversight applies.
## The permission-budget formula [#the-permission-budget-formula]
Treat permissions as a budget, not a setup convenience. Assign a simple consequence weight:
| Permission | Weight |
| ---------------------------------------------- | -----: |
| Public market or blockchain data | 0 |
| Read account balances/history | 1 |
| Place/cancel spot orders | 4 |
| Trade derivatives or use margin | 6 |
| Transfer between account sub-balances | 7 |
| Withdraw assets or sign arbitrary transactions | 10 |
Calculate `authority exposure = sum of enabled permission weights × number of credentialed accounts`.
**Worked hypothetical:** A portfolio analyst has read access to three exchanges: `1 × 3 = 3`. A bot has read and spot-trade access to two exchanges: `(1 + 4) × 2 = 10`. A wallet agent can read and sign arbitrary transactions for one wallet: `1 + 10 = 11`.
This is not a probability of loss or an industry security score. It is a forcing function: the research product with exposure 3 should not be casually treated as equivalent to systems scoring 10 or 11. Any withdrawal or arbitrary-signing permission deserves a separate threat review regardless of the total.
## Decision table: which one fits the job? [#decision-table-which-one-fits-the-job]
| You need | Appropriate starting point | Why |
| ---------------------------------------------- | --------------------------------------- | -------------------------------------------------------- |
| A cited token or protocol investigation | Research agent | Evidence can be reviewed before any decision |
| A daily portfolio briefing | Portfolio-aware research agent | Context and schedule matter; execution does not |
| A notification when a condition changes | Alerting agent or exchange alert | Observation is enough |
| Mechanical execution of a fully specified rule | Trading bot, after testing and controls | The objective is order automation |
| Autonomous interpretation and execution | Highest scrutiny; often unnecessary | Combines model, market, permission, and operational risk |
If your pain is “I do not have time to investigate,” adding trade authority does not solve the research problem. If your pain is “I cannot execute a tested rule consistently,” a conversational research agent may not solve the execution problem.
## A pre-connection checklist [#a-pre-connection-checklist]
* [ ] Write the job in one sentence.
* [ ] List every data source the product needs.
* [ ] Confirm whether the vendor is a researcher, software provider, adviser, broker, exchange, or some combination.
* [ ] Read the exchange or wallet permission screen yourself.
* [ ] Disable trading, transfers, and withdrawals unless the job strictly requires them.
* [ ] Never share a seed phrase or private key with a portfolio tracker.
* [ ] Use a distinct credential for each service.
* [ ] Enable MFA and IP restrictions where supported.
* [ ] Document position, order, slippage, and loss limits for any bot.
* [ ] Test failure cases, not only profitable historical periods.
* [ ] Know how to revoke the key and stop activity before connecting it.
* [ ] Reconcile account activity after connection and review it periodically.
For a balance-only workflow, follow the provider-specific scope details in [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
## Limitations and counterevidence [#limitations-and-counterevidence]
Research-only does not mean risk-free. Read access exposes sensitive financial information, and a persuasive wrong answer can still influence a costly human decision. Conversely, automation is not inherently reckless: carefully specified and supervised systems can reduce manual execution errors. The risk depends on design, permissions, testing, market conditions, and operator discipline.
Backtests are not live evidence. Even an honest backtest may omit latency, liquidity, fees, rejected orders, delistings, survivorship bias, or changing correlations. A profitable history does not establish future performance.
Regulatory status also depends on jurisdiction and what the service actually does. This article is not legal advice; check applicable registrations and obtain professional guidance where necessary.
## The Bottom Line [#the-bottom-line]
An AI crypto agent and a trading bot may analyze the same market, but they occupy different sides of the decision boundary. Classify the product with RACE, inspect its credentials, and minimize authority.
Use a research agent when you want evidence, context, memory, or monitoring. Use a trading bot only when you deliberately want automated order execution and can test and supervise it. Treat a system that both decides and executes as a materially higher-risk product—not simply a smarter version of the same thing.
This article is for research and education, not financial advice. BlockMind does not trade, move funds, or tell you what to buy or sell.
## Sources [#sources]
1. [CFTC: AI Won't Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
2. [FINRA: Risks of auto-trading services offered by unregistered entities](https://www.finra.org/investors/insights/auto-trading-unregistered-entities)
3. [Investor.gov: Artificial intelligence and investment fraud](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/artificial-intelligence-fraud)
4. [Coinbase Exchange: API key permissions](https://help.coinbase.com/en/exchange/managing-my-account/how-to-create-an-api-key)
5. [Kraken: API key permission reference](https://support.kraken.com/articles/360000919966-how-to-create-an-api-key)
6. [OKX: API key permissions and security](https://my.okx.com/docs-v5/)
7. [Binance: Request security and API permissions](https://developers.binance.com/legacy-docs/binance-spot-api-docs/rest-api/request-security)
8. [BlockMind: Trust and security](/docs/security)
---
# AI Crypto Analysis vs X/Twitter: Which Is More Reliable?
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: comparison
Primary keyword: AI crypto analysis vs Twitter
Keywords: AI crypto analysis vs Twitter, crypto Twitter research, X crypto analysis reliability, AI crypto research, verify crypto social media claims
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Neither AI analysis nor X/Twitter is inherently reliable. X is better at exposing you to emerging claims, firsthand reactions, and narrative changes quickly. AI is better at turning a defined claim into a repeatable verification process—if it has current sources and shows its work. The strongest workflow is social media for discovery, primary sources for proof, and AI for synthesis and counterargument.
“AI versus Twitter” is the wrong contest if reliability is the goal. A social feed is a distribution system containing experts, project accounts, anonymous promoters, bots, jokes, and scams. AI is an analysis interface that can use good sources, bad sources, or no current sources at all. Either can mislead you; they fail in different ways.
For the wider product-selection framework, see [the best AI crypto research tools](/blog/best-ai-crypto-research-tools). This article focuses narrowly on social discovery versus structured verification.
## Methodology [#methodology]
Observed July 10, 2026. This is a workflow comparison, not a hands-on benchmark of one model or one X account. It compares the documented capabilities and characteristic failure modes of current web-enabled AI research with public social feeds, using platform policy, regulator evidence, and primary research. Reliability is evaluated claim by claim: identity, source proximity, timestamp, reproducibility, independence, incentives, and counterevidence.
## The comparison at a glance [#the-comparison-at-a-glance]
| Research task | X/Twitter | AI with current tools |
| -------------------------------------- | ------------------------------------------- | ----------------------------------------------------------------- |
| Notice a new narrative | Often excellent and early | Depends on retrieval coverage and prompt |
| Find a project’s public statement | Fast, but impersonators exist | Can retrieve and summarize if the account is resolved correctly |
| Verify a protocol parameter | Weak unless linked to docs or chain | Stronger when directed to primary docs and on-chain evidence |
| Measure consensus | Feed is personalized and non-representative | Can aggregate a defined sample, still vulnerable to sampling bias |
| Compare many sources | Manual and cognitively expensive | Good at structured comparison |
| Detect sarcasm, memes, insider context | Human readers often better | Error-prone |
| Surface contradictions | Depends on who you follow | Good when explicitly asked to red-team |
| Predict price | Unreliable | Unreliable |
| Preserve a research trail | Bookmarks and threads fragment | Can create a cited decision record |
## Where X/Twitter genuinely helps crypto research [#where-xtwitter-genuinely-helps-crypto-research]
### Discovery speed [#discovery-speed]
Developers, researchers, protocols, exchanges, regulators, and market participants publish directly on X. A governance dispute, exploit rumor, listing, technical release, or community reaction can surface there before a retrospective article exists.
This is valuable **discovery**, not automatic verification. The fastest post may be a witness, a mistaken observer, a coordinated promoter, or an account impersonating the source.
### Access to specialist reasoning [#access-to-specialist-reasoning]
Good analysts expose methods, dashboards, contracts, and counterarguments in public. Following a small set of domain experts can teach you which questions matter. A model summary can compress their work, but it should not erase attribution or make several copied posts look like independent evidence.
### Narrative context [#narrative-context]
Markets respond not only to facts but to how participants interpret them. Social attention, repeated phrases, memes, and disagreement can show which narrative is becoming salient. That context is useful even when the narrative is wrong.
Research supports caution rather than a simple “sentiment predicts price” rule. A 2025 open-access study of more than 89,000 Reddit posts found different relationships for Bitcoin and Ethereum and a bidirectional relationship between market dynamics and sentiment—meaning price can influence sentiment as well as the reverse ([Blockchain: Research and Applications](https://doi.org/10.1016/j.bcra.2025.100375)).
## Where a social feed fails [#where-a-social-feed-fails]
### Your feed is not the market [#your-feed-is-not-the-market]
The ranking system, your follows, engagement history, language, and social graph shape what you see. Ten bullish posts in a row may describe a cluster in your feed, not broad investor belief.
### Popularity is not independence [#popularity-is-not-independence]
One claim can be copied across hundreds of accounts. Counting posts without clustering repeated text, links, and account relationships overstates consensus. Engagement can also be purchased or coordinated. X’s own rules prohibit coordinated inauthentic activity and artificial amplification ([X platform manipulation policy](https://help.x.com/en/rules-and-policies/platform-manipulation)).
### Incentives are hidden [#incentives-are-hidden]
An author may hold the asset, receive compensation, seek referral revenue, or front-run followers. A disclosure reduces uncertainty but does not validate the claim. The relevant question is: what evidence would remain if the author’s identity and audience disappeared?
### Fraud uses the same interface as expertise [#fraud-uses-the-same-interface-as-expertise]
The CFTC warns that virtual-currency pump-and-dump schemes use social media and messaging channels to organize hype, especially in thinly traded tokens ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)). The FTC reported that nearly 30% of people who reported losing money to a scam in 2025 said it began on social media, with investment scams representing the largest reported social-media scam losses ([FTC](https://www.ftc.gov/news-events/news/press-releases/2026/04/new-ftc-data-show-people-have-lost-billions-social-media-scams)). Those reports cover more than crypto, but they establish why a social post should never be your sole basis for sending money.
## Where AI analysis helps [#where-ai-analysis-helps]
### It can convert a post into testable claims [#it-can-convert-a-post-into-testable-claims]
Consider: “Whales are accumulating TOKEN before the major upgrade.”
A verification-oriented AI should decompose that into:
1. What contract and chain does TOKEN mean?
2. Which wallets qualify as whales, and who labeled them?
3. What time window defines accumulation?
4. Are transfers purchases, exchange withdrawals, bridge moves, vesting, or internal shuffles?
5. Is the upgrade confirmed in an official release or governance proposal?
6. Did the claim predate the observed move?
That decomposition is more valuable than a bullish/bearish label.
### It can compare unlike evidence consistently [#it-can-compare-unlike-evidence-consistently]
AI can place a project post beside a contract, governance proposal, repository release, audit, market-data series, and counterargument. It can highlight where dates or definitions conflict. The human still has to inspect the decisive sources.
### It can preserve disconfirming evidence [#it-can-preserve-disconfirming-evidence]
Feeds reward novelty and reaction. A research record can keep the original thesis, contrary facts, unknowns, and invalidation conditions visible after the narrative changes. BlockMind’s [Notebook](/docs/memory/notebook) is designed for that durable research context.
## Where AI analysis fails [#where-ai-analysis-fails]
* **Source hallucination:** a model can invent or misstate evidence.
* **Retrieval bias:** search may favor highly linked commentary over a primary document.
* **Freshness mismatch:** a current answer can combine sources from different dates.
* **False aggregation:** ten articles may all repeat one unverified post.
* **Context loss:** sarcasm, irony, deleted context, and quote-post disputes are easy to misread.
* **Authority laundering:** polished prose can make a weak influencer claim sound institutional.
* **Prompt obedience:** asking for a bullish case can produce one even when evidence is poor.
OpenAI explicitly says ChatGPT may produce incorrect facts or fabricated citations and recommends verifying important information, even though search and deep research can improve recency and sourcing ([OpenAI](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)).
## The discovery-to-verification workflow [#the-discovery-to-verification-workflow]
### Step 1: Capture the original claim [#step-1-capture-the-original-claim]
Save the exact URL, author, publication time, wording, and any disclosure. Do not begin with a screenshot; it can omit account identity, edits, replies, or the linked source.
### Step 2: Classify the claim [#step-2-classify-the-claim]
| Claim type | Best verification source |
| ------------------------ | --------------------------------------------------------------------- |
| “The team announced…” | Official site, governance forum, repository, verified project channel |
| “This wallet bought…” | Explorer/indexer plus entity labeling and transaction interpretation |
| “Price broke…” | Named venue, pair, timeframe, and chart data |
| “Users grew…” | Defined on-chain metric or first-party analytics methodology |
| “Partnership confirmed…” | Independent statements from both parties |
| “Audit passed…” | Auditor’s report and exact commit/deployment scope |
| “Everyone is bullish…” | Defined social sample; never infer from one feed |
### Step 3: Ask AI for a source map, not a verdict [#step-3-ask-ai-for-a-source-map-not-a-verdict]
Use this prompt:
> Break this post into factual claims, interpretations, and predictions. For each factual claim, find the earliest primary source, note its date, and quote no more than needed. Identify circular citations and unresolved contradictions. Do not infer a trade.
### Step 4: Check the decisive source yourself [#step-4-check-the-decisive-source-yourself]
Open the contract, proposal, filing, release, or raw dataset. Verify that the AI’s description matches the source and that the source actually addresses the claim.
### Step 5: Compare market reaction with event time [#step-5-compare-market-reaction-with-event-time]
Did the price or social spike happen before or after the claimed catalyst? A true announcement can still be a poor explanation if the move preceded it.
### Step 6: Record uncertainty and monitoring conditions [#step-6-record-uncertainty-and-monitoring-conditions]
Write what remains unknown and what observable event would resolve it. If the question concerns social data, the framework in [how to read crypto social sentiment](/blog/how-to-read-crypto-social-sentiment) helps separate attention, tone, authenticity, and persistence.
## A reliability scorecard for individual claims [#a-reliability-scorecard-for-individual-claims]
Do not assign one credibility score to an account or AI. Score the claim:
* **Identity:** Is the asset, author, and source unambiguous?
* **Proximity:** Is this a primary record or retelling?
* **Timestamp:** Does it predate the event or market move it claims to explain?
* **Reproducibility:** Can another person retrieve the same data and method?
* **Independence:** Are multiple sources genuinely independent?
* **Incentives:** Are holdings, payment, referrals, or conflicts disclosed?
* **Counterevidence:** What credible observation contradicts it?
* **Scope:** Does the conclusion go beyond the evidence?
An answer can be “unresolved.” That is a feature of honest research.
## Limitations and counterevidence [#limitations-and-counterevidence]
X can sometimes be more reliable than a generated report: a signed project announcement or an expert linking directly to raw evidence may be the primary source. AI can also make research worse when it summarizes away caveats or blends independent and copied sources.
Conversely, not every anonymous account is wrong, and not every official account is complete or unbiased. Reliability comes from the evidence chain, not the interface or follower count.
## The Bottom Line [#the-bottom-line]
Use X/Twitter as a radar, not a verdict. Use AI as a structured investigator, not an oracle. For any material claim, preserve the original post, trace it to primary evidence, check timestamps and incentives, seek counterevidence, and record what remains unknown.
That verification boundary also explains [why AI should support rather than issue crypto buy-or-sell decisions](/blog/can-ai-tell-you-when-to-buy-or-sell-crypto).
BlockMind’s agent can research the web and X/Twitter alongside market, on-chain, and portfolio context, but it remains research—not financial advice. It never tells you what to buy or sell and cannot touch funds.
This article is research, not financial advice.
## Sources [#sources]
* [CFTC — Beware Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
* [FTC — New Data Show People Have Lost Billions to Social Media Scams](https://www.ftc.gov/news-events/news/press-releases/2026/04/new-ftc-data-show-people-have-lost-billions-social-media-scams)
* [X — Platform Manipulation and Spam Policy](https://help.x.com/en/rules-and-policies/platform-manipulation)
* [Blockchain: Research and Applications — Virtual Influence, Real Impact](https://doi.org/10.1016/j.bcra.2025.100375)
* [OpenAI — Does ChatGPT Tell the Truth?](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)
---
# AI Crypto Analyst vs ChatGPT: When a Dedicated Agent Still Wins
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: comparison
Primary keyword: AI crypto analyst vs ChatGPT
Keywords: AI crypto analyst, ChatGPT crypto, crypto research tools
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** ChatGPT is already a capable research system. It can search the web, produce cited deep-research reports, remember useful context, connect to apps, and run scheduled monitoring tasks. A dedicated AI crypto analyst wins when you want those pieces pre-wired to live portfolio data, market-specific sources, repeatable expert review, and a daily operating rhythm with less setup.
***
The lazy version of this comparison says ChatGPT is stale, forgetful, and unable to monitor anything. That is no longer true.
ChatGPT can search the current web, run multi-step deep research, use connected apps, reference past conversations, and perform recurring tasks. With the right setup, it can become a serious part of an investor's research process.
The useful question in 2026 is narrower: when is a general AI platform enough, and when is a dedicated crypto analyst worth paying for?
## Methodology [#methodology]
This comparison was reviewed on July 10, 2026. ChatGPT capabilities are taken from current first-party OpenAI documentation linked beside each claim. The dedicated-agent side describes the workflow category and then uses BlockMind's documented product behavior as a concrete example. This is a public-evidence comparison, not a controlled output benchmark or a claim that every specialist product has the same features. Availability, plan limits, and product behavior can change, so verify current terms before choosing a tool.
For a wider product set and the criteria used to choose among research workflows, start with the [AI crypto research tools guide](/blog/best-ai-crypto-research-tools).
## AI crypto analyst vs ChatGPT at a glance [#ai-crypto-analyst-vs-chatgpt-at-a-glance]
| Question | ChatGPT | Dedicated AI crypto analyst |
| ----------------------- | --------------------------------------------------------------- | --------------------------------------------------------------------- |
| General research | Excellent across almost any topic | Strong within its market scope |
| Current web information | Search and deep research with sources | Depends on the agent's connected data and browser tools |
| Portfolio context | Can use information you provide or connect through an app | Usually built around a connected portfolio |
| Long-term context | Memory, chat history, projects, and files | Purpose-built notebook and investment history |
| Monitoring | Scheduled and monitoring tasks are available on eligible plans | Market monitoring is part of the product's default job |
| Specialist data | Available through supported apps or custom integrations | Typically pre-wired to market, on-chain, derivatives, and social data |
| Decision process | You design the prompt or assistant | Market-specific reports and expert review are already defined |
| Setup | Low for one-off research, higher for a durable investing system | Usually ready after account and portfolio setup |
| Breadth | Broad across topics and industries | Deliberately narrower |
ChatGPT wins on versatility. A dedicated agent should win on defaults, continuity, and market-specific execution. If it does not, the specialization is not earning its price.
## What ChatGPT can actually do now [#what-chatgpt-can-actually-do-now]
### Search the current web [#search-the-current-web]
[ChatGPT Search](https://openai.com/index/introducing-chatgpt-search/) can retrieve timely information and return links to sources. It is no longer accurate to say every answer is limited to a static training cutoff.
Search quality still depends on the query and the sources found. Current does not automatically mean correct, and a linked page can be weak evidence. That caveat applies to specialized agents too.
### Produce deep, cited research [#produce-deep-cited-research]
[Deep research in ChatGPT](https://help.openai.com/en/articles/10500283-deep-research-daq) can plan a multi-step investigation, search the public web or selected sites, use uploaded files and connected apps, and produce a structured report with citations.
For a one-off question such as "compare the token economics of these three protocols," that can be exactly the right tool. It is broad, controllable, and good at synthesis.
### Remember useful context [#remember-useful-context]
[ChatGPT Memory](https://help.openai.com/en/articles/8590148-memory-in-chatgpt) can use saved memories and past conversations to personalize later replies. [Projects](https://help.openai.com/en/articles/10169521-projects-in-chatgpt) can also organize related chats, files, and instructions around a long-running objective.
This is real continuity, but it is not the same as a purpose-built investment ledger. OpenAI explicitly notes that memory does not retain every detail, and controls and availability vary. Important theses, levels, and portfolio facts should live in a structured source you can inspect rather than depending on conversational recall alone.
### Connect to external data [#connect-to-external-data]
[Apps in ChatGPT](https://help.openai.com/en/articles/11487775-connectors-in-chatgpt) can provide access to external tools and information. Deep research can use connected sources, and eligible workspaces can support custom apps built with MCP when their settings allow it.
That means a team can connect proprietary market or portfolio data. The limitation is not possibility. It is setup, availability, maintenance, and whether the exact data source you need has an appropriate app.
### Run scheduled monitoring [#run-scheduled-monitoring]
[Scheduled Tasks in ChatGPT](https://help.openai.com/en/articles/10291617-tasks-in-chatgpt) support one-off, recurring, and monitoring tasks. A task can periodically check for a change and notify the user when something meaningful happens.
This removes another outdated criticism. ChatGPT can work between prompts. The practical limits still matter: task availability and active-task limits depend on the plan, runs cannot be more frequent than hourly, unattended tasks can pause, and a task created inside a project cannot access that project's files.
## Where a dedicated AI crypto analyst still wins [#where-a-dedicated-ai-crypto-analyst-still-wins]
### Portfolio context is the starting point [#portfolio-context-is-the-starting-point]
In a dedicated product, portfolio context is not something you paste into a prompt or wire up later. Read-only wallet and exchange connections make holdings available to the research system by default.
That changes the first question. Instead of "what happened to ETH?" the analyst can begin with "what does this ETH move mean for your exposure, your saved thesis, and the rest of your portfolio?"
### The data stack is already selected [#the-data-stack-is-already-selected]
General research tools must find or connect sources for each job. A specialist product can arrive with market prices, on-chain data, derivatives, social signals, news, portfolio data, and technical tools already available.
This does not guarantee accuracy. It does remove repeated setup and makes the research process more consistent from one asset to the next.
### The research shape is repeatable [#the-research-shape-is-repeatable]
A blank prompt is flexible, but flexibility creates variance. One token gets a full investigation; the next gets a quick summary because the user forgot three checks.
A dedicated analyst can enforce a consistent shape:
* Explore the market for candidates
* Analyze one idea across defined evidence categories
* Ask several experts to challenge the conclusion
* Track the saved thesis after the report is complete
The user can still ask an open question. The default path simply makes omissions less likely.
### Monitoring begins without designing a task system [#monitoring-begins-without-designing-a-task-system]
ChatGPT Tasks can monitor changes, but the user must decide what to schedule, provide accessible context, and manage the task limits.
A dedicated investing agent can begin with sensible portfolio and market checks already active, then let the user add durable rules in plain language. The value is not that scheduling is exclusive. It is that monitoring is a core product contract.
### The daily brief is a product, not a prompt [#the-daily-brief-is-a-product-not-a-prompt]
You can ask ChatGPT to prepare a recurring market summary. A specialist agent can ground the brief in connected holdings, saved verdicts, watchlists, and prior research without rebuilding that package each time.
By default, BlockMind publishes its [Morning Brief](/blog/crypto-morning-brief) to the dashboard and sends it by email at about 8:00 in the user's timezone. That delivery and context are part of the agent's standard job.
## Where ChatGPT is the better choice [#where-chatgpt-is-the-better-choice]
Specialization is not always an advantage.
Choose ChatGPT when:
* Your research crosses markets, industries, legal questions, technology, and many other domains
* You want a one-off cited report without connecting a portfolio
* You already maintain clean investment data in files or compatible apps
* You enjoy designing prompts, projects, tasks, and custom integrations
* You need a flexible general assistant for far more than investing
* You want to pressure-test the specialist agent with an independent second view
ChatGPT can also be the better writing and explanation partner. A narrow agent may have better market defaults while a general model is better at changing format, teaching a concept, or exploring an unusual analogy.
## Where a dedicated agent is the better choice [#where-a-dedicated-agent-is-the-better-choice]
Choose a specialist when:
* You want the portfolio connected read-only and available in every analysis
* You want market-specific sources without building the data stack
* You repeatedly research assets with the same due-diligence checklist
* You want scheduled portfolio monitoring from day one
* You want a daily brief tied to holdings and saved theses
* You want investment criteria stored for reusable expert review
* You do not want to maintain prompts and integrations as a side project
The time saved should come from operations, not just faster prose.
## Can you build your own crypto analyst with ChatGPT? [#can-you-build-your-own-crypto-analyst-with-chatgpt]
Yes, within the capabilities and plan limits available to you.
A serious setup could combine:
1. A Project for research instructions and durable files
2. Memory for stable preferences
3. Deep research for cited investigations
4. Apps or a custom MCP integration for portfolio and market data
5. Scheduled Tasks for recurring checks
6. A structured external notebook for theses, sources, and decisions
That can be powerful. It is also a small systems-integration project. You own data permissions, prompt design, source selection, task behavior, maintenance, and the handoffs between them.
There is nothing wrong with building it. The buy-versus-build question is whether maintaining the analyst is how you want to spend your research time.
## The hallucination problem applies to both [#the-hallucination-problem-applies-to-both]
Specialized branding does not remove model error. Both general and dedicated AI systems can:
* Misread a source
* Confuse similarly named assets
* Overstate the cause of a price move
* Fill a data gap with a plausible claim
* Cite a source that does not fully support the conclusion
* Sound more certain than the evidence allows
The correct defense is visible sources, repeatable checks, contradiction seeking, and human review. Never trust a claim merely because the system knows your portfolio.
## How BlockMind differs from ChatGPT [#how-blockmind-differs-from-chatgpt]
BlockMind gives Pro members one named personal AI investing agent with its own workspace. It is pre-wired for crypto, stocks, and commodities, with connected portfolio context and market-specific research tools.
The product organizes work as Explore, Analyze, Verdicts, and Track. The agent prepares a Morning Brief, runs periodic monitoring, stores research and preferences in a Notebook, and can ask several experts to review an analysis.
The trust boundary is balance-only inside BlockMind. Wallet and exchange connections let the agent see positions, but it cannot trade, withdraw, or move funds even if you told it to. Provider key scopes vary; the [connection guide](/docs/getting-started/connect-your-portfolio) explains the Firi exception.
See [current BlockMind plans, introductory terms, and the card requirement](/docs/overview/plans-and-pricing) on the canonical pricing page. The free classic platform remains available for tracking, watchlists, charts, and market indicators without the personal agent.
## Frequently Asked Questions [#frequently-asked-questions]
### Is ChatGPT good for crypto research? [#is-chatgpt-good-for-crypto-research]
Yes. Search and deep research make it strong for current, cited investigations, especially when you provide a precise question and trustworthy sources. You still need to verify material claims.
### Does ChatGPT remember my portfolio? [#does-chatgpt-remember-my-portfolio]
It can remember useful context, use files, and connect to external data through apps. It does not automatically have a complete, live portfolio unless you provide or connect one. Memory is also not guaranteed to retain every detail.
### Can ChatGPT monitor crypto prices for me? [#can-chatgpt-monitor-crypto-prices-for-me]
Eligible plans support scheduled and monitoring tasks. Runs cannot occur more than once per hour, limits vary by plan, and task context has constraints. Use exchange-native alerts when timing is critical.
### Is BlockMind just a custom ChatGPT? [#is-blockmind-just-a-custom-chatgpt]
No. The distinction is not merely the model. BlockMind provisions a dedicated workspace, connects portfolio data for balance-only use, supplies market tools, schedules monitoring and Morning Briefs, stores a structured Notebook, and provides the Explore, Analyze, Verdicts, Track research journey.
### Should I use both? [#should-i-use-both]
That can be sensible. Use the dedicated agent for recurring portfolio work and ChatGPT as an independent general researcher or second opinion. Agreement is useful; disagreement is often more useful because it exposes assumptions to check.
## The Bottom Line [#the-bottom-line]
ChatGPT has closed many of the gaps that once justified a specialist by default. It can search, research, remember, connect, and monitor. Any dedicated AI crypto analyst should be compared against that current product, not the ChatGPT of several years ago.
A specialist still earns its place when it removes the integration work and maintains the research habit for you. Choose ChatGPT for breadth and flexibility. Choose a dedicated agent for portfolio-native defaults, repeatable market research, and continuity with less setup.
## Learn More [#learn-more]
* [What Is an AI Crypto Agent?](/blog/what-is-an-ai-crypto-agent)
* [AI Portfolio Monitoring](/blog/ai-portfolio-monitoring-crypto)
* [BlockMind vs CoinStats](/blog/blockmind-vs-coinstats)
* [Crypto Morning Brief](/blog/crypto-morning-brief)
## Sources [#sources]
1. [Introducing ChatGPT Search](https://openai.com/index/introducing-chatgpt-search/)
2. [Deep research in ChatGPT](https://help.openai.com/en/articles/10500283-deep-research-daq)
3. [Memory FAQ](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)
4. [Apps in ChatGPT](https://help.openai.com/en/articles/11487775-connectors-in-chatgpt)
5. [Scheduled Tasks in ChatGPT](https://help.openai.com/en/articles/10291617-tasks-in-chatgpt)
6. [Projects in ChatGPT](https://help.openai.com/en/articles/10169521-projects-in-chatgpt)
7. [What is BlockMind?](/docs/overview/what-is-blockmind)
8. [The BlockMind research journey](/docs/capabilities/research-journey)
---
# AI Crypto Tool vs Price Bot: What Is the Difference?
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: comparison
Primary keyword: AI crypto tool vs price bot
Keywords: AI crypto tool vs price bot, crypto price bot, AI crypto analysis tool, crypto alert bot, crypto research agent
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A price bot performs narrow, deterministic jobs: retrieve a quote, watch a threshold, or forward a market event. An AI crypto tool performs interpretive jobs: combine sources, explain possible causes, relate a change to a portfolio, and update a research record. Use a price bot when speed and predictable rules matter. Use an AI tool when the question needs context. Use neither as an unquestioned trade signal.
The labels are messy. Some products called “AI bots” are simple threshold automations; some price apps now summarize news; some agents can schedule alerts. The reliable way to compare them is by the work they perform and the evidence they expose.
Use [the best AI crypto research tools](/blog/best-ai-crypto-research-tools) for the broader category comparison; the distinction here is between deterministic price-event automation and probabilistic interpretation.
## Methodology [#methodology]
Observed July 10, 2026. This compares two tool architectures rather than ranking named products. The matrix is based on the jobs each system performs, official market-data documentation, and regulator guidance about AI and trading-bot claims. It is not a hands-on latency, accuracy, or return test. Individual products can combine both architectures, so evaluate their actual data sources, permissions, rules, and evidence.
## Price bot and AI crypto tool at a glance [#price-bot-and-ai-crypto-tool-at-a-glance]
| Dimension | Price bot | AI crypto tool |
| -------------- | ----------------------------------- | -------------------------------------------------------------------------- |
| Core job | Retrieve or compare a number | Interpret several pieces of evidence |
| Typical input | Price feed and rule | Market, on-chain, news, social, portfolio, and prompt context |
| Typical output | Quote or notification | Explanation, report, scenario, or prioritized alert |
| Behavior | Mostly deterministic | Probabilistic and language-driven |
| Best question | “Did BTC/USD cross X on venue Y?” | “What changed, what evidence explains it, and which holdings are exposed?” |
| Latency | Often lower | Usually higher because retrieval and reasoning take time |
| Auditability | Rule is easy to inspect | Depends on citations, tool logs, and stated method |
| Main risk | Bad feed, wrong pair, noisy trigger | Hallucination, weak sources, overinterpretation |
| Custody needed | No | No for research |
## What a price bot actually does [#what-a-price-bot-actually-does]
A price bot is an interface around market data and conditional logic. Its useful primitives are simple:
* return a current quote for a specified asset and pair;
* calculate a percentage change over a fixed window;
* alert when a value crosses, closes above, or remains beyond a threshold;
* forward volume, funding, liquidation, or order-book events;
* render a basic chart from structured data.
The simplicity is a strength. “Send an alert when ETH/USD on venue A trades at or below X” can be tested mechanically. The bot should not need to invent an explanation.
The underlying data still matters. Coinbase’s Exchange WebSocket is described as a real-time public feed for orders and trades, yet its documentation warns that messages can be dropped or arrive out of order and explains how sequence numbers reveal gaps ([Coinbase](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)). A competent alert system must recover from that condition rather than silently treating an incomplete stream as the market.
Aggregated APIs trade some latency for broad coverage and normalized data. CoinGecko’s documented simple-price endpoint supports price, market cap, 24-hour volume and change, plus a last-updated timestamp; its listed Pro plans use a 20-second cache/update frequency ([CoinGecko](https://docs.coingecko.com/reference/simple-price)). Neither source is universally “better.” The right choice depends on whether you need venue-specific execution data or a broad market research snapshot.
## What an AI crypto tool adds [#what-an-ai-crypto-tool-adds]
An AI tool becomes useful when the output cannot be defined as one formula or threshold.
### It can route the question to several sources [#it-can-route-the-question-to-several-sources]
“Why did my portfolio fall today?” may require:
* normalized holdings and prices;
* asset and narrative exposure;
* market-wide moves;
* project-specific news;
* on-chain events;
* derivatives or social context;
* comparison with the previous portfolio state.
The model’s job is to organize that evidence and explain which factors are confirmed, plausible, or unknown. The model is not itself the data source. Read [how AI accesses real-time crypto market data](/blog/how-ai-accesses-real-time-crypto-market-data) for the complete event-to-answer path.
### It can judge relevance [#it-can-judge-relevance]
A 9% move in an asset you do not hold may be noise. A smaller change in a concentrated position, a collateral asset, or a shared protocol dependency may matter more. A portfolio-aware AI can rank alerts by exposure instead of raw magnitude.
### It can preserve research context [#it-can-preserve-research-context]
A threshold bot knows that price crossed X. A persistent research agent can also know why you chose X, which thesis it relates to, and what other evidence was supposed to accompany the event. That context turns a notification into a review prompt rather than a reflex.
### It can explain uncertainty [#it-can-explain-uncertainty]
A trustworthy answer can say: “The price move is confirmed; these two events occurred in the same window; causation is unresolved.” That is more honest than attaching the most popular headline to every move.
## Where a price bot is the better tool [#where-a-price-bot-is-the-better-tool]
Choose a deterministic bot when:
* the trigger can be written as a precise rule;
* low latency matters more than interpretation;
* you want consistent behavior every time;
* a false explanation would be worse than no explanation;
* you can define the venue, pair, time window, and reset behavior;
* you need a lightweight notification rather than a report.
Examples:
* “Alert once if BTC/USD closes below X on a 4-hour candle.”
* “Notify me if this exchange’s spread remains above Y for five minutes.”
* “Report the daily price at 08:00 UTC with source and timestamp.”
The details prevent alert noise. “Price touches X” behaves differently from “candle closes beyond X.” A one-shot trigger differs from an alert that fires on every tick. Crossing direction, cooldown, source failure, and re-entry rules should be explicit.
## Where an AI crypto tool is the better tool [#where-an-ai-crypto-tool-is-the-better-tool]
Choose AI-assisted analysis when:
* the question spans several data types;
* source quality needs comparison;
* you want an explanation in portfolio context;
* the alert depends on materiality, not just magnitude;
* you want counterevidence and unresolved questions;
* the result should update a thesis or research journal.
Examples:
* “Explain which holdings drove today’s drawdown and separate market-wide from asset-specific evidence.”
* “Check whether this token’s social spike is accompanied by liquidity and on-chain activity.”
* “Review the governance proposal and identify the assumptions in my saved thesis that it changes.”
* “Classify this whale transfer before alerting me; do not call it a sale without exchange or swap evidence.”
BlockMind’s [monitoring and alerts](/docs/automations/monitoring-and-alerts) are designed around this contextual job, while its agent remains unable to trade or move funds.
## The best architecture uses both [#the-best-architecture-uses-both]
A robust workflow separates detection from interpretation:
1. **A deterministic monitor detects an event.**
2. **A validation step confirms the source is healthy and the trigger is real.**
3. **The system retrieves relevant context.**
4. **AI summarizes facts, hypotheses, portfolio relevance, and unknowns.**
5. **The user decides whether anything should happen.**
This separation contains failure. If the AI explanation fails, the underlying threshold remains auditable. If the feed fails, the system should report degraded data rather than invite the AI to guess.
## Seven questions to ask before choosing a tool [#seven-questions-to-ask-before-choosing-a-tool]
### 1. What source defines the price? [#1-what-source-defines-the-price]
Require the venue or aggregation method, pair, currency, price type, and timestamp. “ETH price” is incomplete.
### 2. What happens when data is missing? [#2-what-happens-when-data-is-missing]
Look for gap detection, stale-data labels, retries, fallback rules, and explicit failure states. Silent substitution is dangerous.
### 3. Is the output a fact or an inference? [#3-is-the-output-a-fact-or-an-inference]
“Price crossed X” is observable. “Whales caused the move” is an interpretation requiring separate evidence.
### 4. Can you inspect and reproduce the alert? [#4-can-you-inspect-and-reproduce-the-alert]
The rule, timeframe, cooldown, and historical event should be visible. For AI analysis, sources and assumptions should be reviewable.
### 5. Does it know the relevant portfolio safely? [#5-does-it-know-the-relevant-portfolio-safely]
Research needs only read-only data. Never provide seed phrases, private keys, withdrawal rights, or signing authority. The current BlockMind connection scope is documented in [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
### 6. Does it promise outcomes? [#6-does-it-promise-outcomes]
The CFTC warns that promoters use AI and trading-bot language to market unrealistic or guaranteed returns; AI cannot predict the future or sudden market changes ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)). A tool that claims a perfect win rate fails the trust test before its feature list matters.
### 7. Can you export or preserve the evidence? [#7-can-you-export-or-preserve-the-evidence]
A useful alert should retain the triggering value, timestamp, source, and relevant context. Otherwise you cannot audit whether it helped.
## Common category mistakes [#common-category-mistakes]
### Calling summarization “real-time analysis” [#calling-summarization-real-time-analysis]
An AI can summarize a cached page fluently. That does not make the input live. Check the source timestamp.
### Calling a threshold “AI” [#calling-a-threshold-ai]
A fixed rule is not more valuable because the interface uses conversational language. Evaluate the rule and feed.
### Asking AI to do a feed’s job [#asking-ai-to-do-a-feeds-job]
For exact time-sensitive quotes, structured market data is the source of truth. AI should explain the number, not recall it.
### Asking a price bot to explain causation [#asking-a-price-bot-to-explain-causation]
A threshold crossing says nothing about why it occurred. Adding the nearest headline is not causal analysis.
### Treating either tool as financial advice [#treating-either-tool-as-financial-advice]
An accurate alert can still be irrelevant to your horizon, taxes, liabilities, and risk. A good explanation can still be wrong.
## Limitations and counterevidence [#limitations-and-counterevidence]
The boundary is not permanent. Modern price products can add news and portfolio context, while AI agents can execute deterministic scheduled checks. The categories overlap at the interface.
The underlying distinction remains useful: deterministic detection and probabilistic interpretation have different failure modes. Combining them does not eliminate those risks. It requires more careful logging, source health checks, and uncertainty labels.
## The Bottom Line [#the-bottom-line]
Use a price bot for precise, repeatable detection. Use an AI crypto tool for sourced interpretation and portfolio relevance. The strongest setup lets a deterministic rule wake the analysis layer, keeps facts separate from hypotheses, and leaves every financial decision with you.
For another boundary that is often blurred in product names, compare an [AI crypto agent with a trading bot](/blog/ai-crypto-agent-vs-trading-bot).
BlockMind is a research agent, not a trading bot. It can analyze and monitor read-only portfolio context, but it never tells you what to buy or sell and cannot touch funds. This is research, not financial advice.
## Sources [#sources]
* [Coinbase Developer Documentation — Exchange WebSocket Overview](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)
* [CoinGecko API — Simple Price Endpoint](https://docs.coingecko.com/reference/simple-price)
* [CFTC — AI Won’t Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
* [NIST — Generative AI Risk Management Profile](https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=958388)
---
# AI Portfolio Monitoring: An Agent That Watches Your Crypto While You Sleep
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: problem-solving
Primary keyword: AI crypto portfolio monitoring
Keywords: AI portfolio monitoring, crypto portfolio alerts, crypto risk monitoring
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** AI portfolio monitoring is useful when it watches the reasons behind your positions, not just their prices. A good agent checks holdings and market risks on a schedule, compares new evidence with your thesis, and explains only the changes that matter. It should never need permission to trade or move funds.
***
Portfolio tracking tells you what your assets are worth when you open the app. Portfolio monitoring asks a harder question while you are away: did anything change that deserves your attention?
That could be a sharp move, but it could also be an exploit, a token unlock, a change in holder behavior, a missed catalyst, or new evidence that weakens a saved thesis. The best signal is not always a price notification.
## What is AI portfolio monitoring? [#what-is-ai-portfolio-monitoring]
AI portfolio monitoring is the use of an agent to review holdings, tracked assets, and broader market conditions on a schedule, then rank and explain material changes.
The process has four parts:
1. **Observe:** collect current portfolio, market, on-chain, and news data.
2. **Compare:** check it against previous values, alert rules, and open research.
3. **Judge relevance:** decide what could matter to this portfolio.
4. **Report:** send an alert or include the change in the next briefing.
Traditional alerts handle the first comparison well. An agent becomes useful in the third step, where several facts need context.
## Portfolio tracking vs portfolio monitoring [#portfolio-tracking-vs-portfolio-monitoring]
| Portfolio tracking | Portfolio monitoring |
| ------------------------------------- | ------------------------------------------ |
| Shows balances and performance | Checks for material change on a schedule |
| Starts when you open the product | Runs between visits |
| Treats every user interface similarly | Uses your holdings, watchlists, and theses |
| Reports values | Explains why a change may matter |
| Usually ends at a notification | Can continue into research and follow-up |
You still need accurate tracking underneath. Monitoring cannot fix bad balances or incomplete transaction data. It adds interpretation and continuity on top.
## What should an AI agent monitor? [#what-should-an-ai-agent-monitor]
### Large moves in your holdings [#large-moves-in-your-holdings]
Price remains the fastest clue that the market knows something. A monitoring pass can identify outliers, compare them with sector and market moves, and look for a plausible driver.
The explanation matters. "Asset down 9%" is a notification. "Asset down 9% while its sector is flat, after a treasury transfer to an exchange" is a research lead.
### Assets you follow but do not own [#assets-you-follow-but-do-not-own]
Watchlists are where future positions and rejected ideas live. Monitoring can flag a target price, catalyst, governance proposal, earnings release, token unlock, or change in the evidence that kept you on the sidelines.
This is especially useful after an analysis ends with "wait." The agent can preserve what you are waiting for.
### Market-wide risks [#market-wide-risks]
Some events affect many positions at once:
* Stablecoin depegs
* Protocol exploits and bridge halts
* Exchange or custody incidents
* Sudden leverage and liquidation stress
* Large changes in Bitcoin dominance
* Macro releases that alter risk appetite
One market-level warning can be more useful than ten separate asset alerts.
### Thesis invalidations [#thesis-invalidations]
The strongest monitoring instruction is not "tell me when the price moves." It is "tell me when the reason I own this stops being true."
That requires a written thesis with observable conditions. Examples:
* User growth remains above a threshold
* A product launch happens before a deadline
* Token supply stays below a planned unlock level
* Revenue growth continues while valuation stays within a range
* A chart level holds after a catalyst
An agent cannot monitor a reason you never recorded.
### Portfolio concentration and correlation [#portfolio-concentration-and-correlation]
Five token names can still be one trade if they depend on the same narrative, chain, liquidity source, or macro condition. A portfolio-aware agent can flag when exposure becomes more concentrated than the asset count suggests.
## Why simple price alerts are not enough [#why-simple-price-alerts-are-not-enough]
Price alerts are precise, fast, and useful. They are also narrow.
An 8% move may be ordinary volatility for one asset and a thesis break for another. A price can remain flat while the underlying risk changes sharply. A token unlock can increase future selling pressure before the market reacts. An exploit elsewhere can expose a shared dependency.
AI monitoring should not replace exact alerts. It should put them in context and catch changes that do not fit a single numeric rule.
## What "always-on" should mean [#what-always-on-should-mean]
"Always-on" is often misunderstood as real-time market surveillance. Most personal agents work through scheduled checks.
That is appropriate for research and portfolio awareness, but not for execution. A condition can occur between passes, and an alert may arrive later. If your strategy depends on seconds or minutes, use exchange-native alerts and dedicated trading infrastructure.
For a research agent, always-on should mean:
* It keeps the monitoring instruction after the conversation ends
* It checks again without another prompt
* It records what it found
* It reports through a reliable channel
* It stops or changes the rule when you ask
Honest frequency is more useful than vague claims of real-time intelligence.
## How to write a monitoring instruction [#how-to-write-a-monitoring-instruction]
Give the agent a condition, context, and response.
Weak:
> Watch ETH.
Better:
> Check ETH during each monitoring pass. Alert me if it moves 8% in 24 hours, if ETH/BTC falls below 0.03, or if major staking withdrawals change the liquidity picture. Explain the likely cause and relate it to my current ETH exposure.
The second instruction defines what counts, what context matters, and what a useful response contains.
For thesis monitoring:
> Track my saved SOL analysis. Flag evidence that changes the validator, developer-activity, or institutional-flow case. Do not repeat the original thesis unless new evidence strengthens or weakens it.
This prevents the agent from generating activity without information.
## Where alerts should arrive [#where-alerts-should-arrive]
The right channel depends on urgency.
* **In-app:** durable record, details, and follow-up research
* **Email:** useful for material events you should not miss
* **Morning Brief:** lower-urgency changes that belong in daily context
Optional companion channels can be useful when enabled, but they should not be assumed. The alert should always remain visible in the app.
## Security: context without custody [#security-context-without-custody]
Monitoring needs visibility, not control.
Wallet addresses are public. Exchange integrations can use API keys restricted to read-only permissions. The agent should never receive a seed phrase, private key, withdrawal permission, or trading permission just to monitor a portfolio.
With BlockMind, wallet and exchange connections are used only to read balances and positions. The agent cannot trade, withdraw, or move funds even if you told it to.
That separation protects the user and improves the product's incentives. The agent can surface evidence consistent with waiting without an execution system standing by to act.
## How BlockMind monitors a portfolio [#how-blockmind-monitors-a-portfolio]
BlockMind Pro includes a personal AI investing agent with its own workspace. The agent runs scheduled monitoring passes over connected portfolios, tracked assets, and broader market risks.
It can:
* Flag material moves in holdings and watchlists
* Check major news and market-wide risk events
* Evaluate persistent price and ratio alerts written in plain language
* Keep recurring checks you ask it to run
* Connect a saved analysis or expert verdict to future monitoring
* Summarize lower-urgency changes in the next [Morning Brief](/blog/crypto-morning-brief)
Alerts land in the app and can be sent by email. The monitoring is periodic, not tick-level. It is designed to preserve context and reduce blind spots, not to time entries or execute trades.
The free classic platform still tracks portfolios and watchlists. The scheduled personal agent comes with Pro.
## Frequently Asked Questions [#frequently-asked-questions]
### Can AI monitor a crypto wallet? [#can-ai-monitor-a-crypto-wallet]
Yes. Public wallet addresses can be monitored without custody, and exchange credentials can be used for balance-only monitoring. Use provider-scoped read-only keys wherever available; Firi does not offer that scope, so BlockMind enforces the restriction in the application. The agent should never need a seed phrase or withdrawal access.
### Is AI portfolio monitoring real time? [#is-ai-portfolio-monitoring-real-time]
Not necessarily. Many agents run periodic checks. Ask for the actual frequency and use exchange-native systems when seconds matter.
### Can an agent alert me when a token reaches a price? [#can-an-agent-alert-me-when-a-token-reaches-a-price]
Yes, if the product supports persistent alert rules. A useful alert should include portfolio context and a likely explanation, not only the crossed number.
### Can an AI agent sell a position automatically? [#can-an-ai-agent-sell-a-position-automatically]
Some trading systems can, but a research agent does not need that authority. BlockMind is read-only and cannot trade, withdraw, or move funds.
### What happens if my BlockMind Pro subscription ends? [#what-happens-if-my-blockmind-pro-subscription-ends]
The agent pauses. Your portfolio, briefs, workspace, and saved context remain available for the agent if you restart it later.
### How much does BlockMind monitoring cost? [#how-much-does-blockmind-monitoring-cost]
Monitoring is part of the Pro agent. See [current plans, introductory terms, and the card requirement](/docs/overview/plans-and-pricing) on the canonical pricing page.
## The Bottom Line [#the-bottom-line]
Good portfolio monitoring is not a louder notification system. It is a maintained link between what you own, why you own it, and what has changed.
Use exact alerts for exact numbers. Use an agent for context, repetition, and thesis drift. Keep every connection read-only, demand honest monitoring frequency, and reserve execution for systems built and permissioned for execution.
## Learn More [#learn-more]
* [Crypto Morning Brief](/blog/crypto-morning-brief)
* [What Is an AI Crypto Agent?](/blog/what-is-an-ai-crypto-agent)
* [AI Crypto Analyst vs ChatGPT](/blog/ai-crypto-analyst-vs-chatgpt)
* [BlockMind vs CoinStats](/blog/blockmind-vs-coinstats)
* [What to Check Before Buying Crypto](/blog/what-to-check-before-buying-crypto)
* [BlockMind monitoring and alerts](/docs/automations/monitoring-and-alerts)
## Sources [#sources]
1. [Coinbase: API key authentication and permissions](https://docs.cdp.coinbase.com/coinbase-app/authentication-authorization/api-key-authentication), accessed July 2026.
2. [Kraken: How to create an API key](https://support.kraken.com/hc/en-us/articles/360000919966-how-to-create-an-api-key), accessed July 2026.
3. [NIST: AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework), accessed July 2026.
---
# AI vs Manual Crypto Research: What to Automate and What to Judge Yourself
Published: 2026-07-03
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: comparison
Primary keyword: AI vs manual crypto research
Keywords: AI crypto research, manual crypto research, crypto research workflow
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** AI is better than you at gathering data, cross-referencing sources, monitoring positions around the clock, and flagging patterns. You are better than AI at judging founder quality, timing narratives, reading incentive design, and sizing positions. Investors who get this split right do deeper research in less time because neither side is doing the other's job.
***
The question is not whether AI can research crypto. With the right tools, it can read a whitepaper, pull holder data, and summarize a week of social sentiment faster than any human. The real question is which parts of the research process you should hand over and which parts you should never hand over.
This post is the honest version of that answer. Not "AI does everything" (it doesn't), and not "nothing beats manual DYOR" (a claim usually made by people who stopped doing manual DYOR months ago because it takes too long).
## Methodology [#methodology]
**Observation date: July 10, 2026.** This is a task-allocation comparison, not a timed contest between one model and one analyst. The framework separates work by auditability, repetition, consequence, and dependence on personal context. The time example below is BlockMind's disclosed planning baseline for a disciplined first pass, not a measured industry average. Claims about AI limitations are bounded by current risk guidance and may change as tools improve; the human retains every portfolio and execution decision throughout.
## What manual research actually costs [#what-manual-research-actually-costs]
Start with an illustrative baseline, not a universal benchmark. BlockMind's checklist allocates **about 2–3 hours** to a disciplined first pass: roughly 30 minutes on the team, 20 on tokenomics, 45 on on-chain analysis, 30 on sentiment, and 20 on technicals before you write a conclusion. The real time varies by asset, available evidence, and depth. We break down that estimate in [How long does it take to research a crypto token properly?](/blog/how-long-to-research-crypto-token).
And that's one token, once. The full process includes team, whitepaper, tokenomics, audits, holders, liquidity, sentiment, on-chain activity, competition, community, technicals, and risk/reward. It is a [12-step checklist](/blog/manual-dyor-checklist-12-steps), and the honest finding from that post applies here too: most people skip half the steps, not because any single step is hard, but because doing all of them consistently is exhausting.
The costs compound in three ways:
* **Fragmentation.** Each step lives in a different tool: a block explorer, an unlock tracker, a social feed, or a charting platform. You spend as much time switching contexts as analyzing.
* **Staleness.** Research is a snapshot. The token you researched three weeks ago has new unlocks, new holders, new narrative. Manual research doesn't update itself.
* **Inconsistency.** You spend three hours on one token and twelve minutes on the next, then treat both conclusions with equal confidence. That's not a research process; that's a mood.
None of this means manual research is bad. It means manual research is *expensive*, and expensive things should be spent where they earn the most.
## What AI does better than you [#what-ai-does-better-than-you]
Four jobs, specifically. These are the mechanical layers of research, where being tireless and systematic beats being smart.
### Data gathering [#data-gathering]
Pulling market data, supply figures, holder distributions, TVL, fee revenue, and unlock schedules is not analysis. It is collection. A human doing this is a slow, error-prone API. AI compresses hours of tab-hopping into minutes, and it doesn't "forget" to check the unlock schedule because it's Friday afternoon.
### Cross-referencing [#cross-referencing]
The most valuable signals in crypto research come from *contradictions between sources*: social hype rising while active users fall, a "community-owned" treasury that on-chain data shows concentrated in three wallets, reported volume that liquidity depth can't support. Humans are bad at this because it requires holding six data sources in your head simultaneously. Machines are built for it.
### Monitoring [#monitoring]
Research decays. The thesis you built in March needs to be stress-tested against April's unlock, May's governance vote, and June's exploit in a competitor. No human re-runs their research on every holding every day. Software can. Monitoring is arguably where automation delivers the most value per hour saved because the alternative usually is not slower monitoring; it is *no monitoring*.
### Pattern flags [#pattern-flags]
Concentrated holders, aggressive vesting cliffs, inflation-funded yield, and wash-like volume are known shapes. Once a red flag has a definition, checking for it is a mechanical task, and mechanical tasks should never be the reason your research took three hours. AI flags the pattern; you decide whether the pattern matters in this case.
## What humans keep [#what-humans-keep]
Now the other side of the ledger: the judgments you should not delegate because they depend on taste, context, and skin in the game.
### Founder quality [#founder-quality]
AI can verify that a founder exists, has a LinkedIn, and shipped previous projects. It cannot tell you whether they're a killer or a tourist. That read comes from watching how they answer hard questions, whether they ship through drawdowns, and whether their past collaborators actually vouch for them. Data narrows the question; judgment answers it.
### Narrative timing [#narrative-timing]
Knowing that a narrative exists is a data problem. Knowing whether you're *early or late* to it is a judgment problem, and it is most of the game. The same token can be a great idea in month two of a narrative and exit liquidity in month eight. No dataset labels which month you're in.
### Incentive design [#incentive-design]
Tokenomics data tells you the allocations and the unlock dates. It doesn't tell you what the people holding those allocations will *do*. Reading incentive design means asking who wins if this succeeds, who gets paid either way, and whose interests quietly diverge from yours. That is game theory applied to specific humans. Our [tokenomics research guide](/blog/how-to-research-crypto-tokenomics) covers the data side; the interpretation stays with you.
### Position sizing and risk [#position-sizing-and-risk]
How much to risk, where your invalidation is, and whether this position improves your portfolio or just adds correlated exposure depend on your goals, timeline, and tolerance for being wrong. No tool should make those decisions, and you should be suspicious of any tool that offers to.
## A practical split of the work [#a-practical-split-of-the-work]
Here's how the division looks in practice, per token:
| Stage | Who does it | What happens |
| ------------- | ----------- | ------------------------------------------------------------------------------------------------------------------------------- |
| 1. Screen | AI | Gather the basics: supply, holders, liquidity, unlocks, obvious red flags. Kill weak candidates fast. |
| 2. Deep pass | AI + you | AI compiles team background, on-chain activity, sentiment, and competition. You read it critically and note what surprises you. |
| 3. Judgment | You | Founder quality, narrative timing, incentive design. Write your thesis and your invalidation in plain language. |
| 4. Decision | You | Size the position or decide not to take it. This step is never automated. |
| 5. Monitoring | AI | Watch the position, the thesis, and the invalidation. Flag changes; you re-judge only when something actually moves. |
The pattern: AI owns steps 1, 2, and 5 (collection, compilation, vigilance). You own steps 3 and 4 (judgment, commitment). Step 2 is shared: the machine assembles, and the human interrogates.
If you're deciding what belongs in the deep pass, our guide on [what to check before buying crypto](/blog/what-to-check-before-buying-crypto) is the checklist to hand over.
## Where BlockMind Fits [#where-blockmind-fits]
BlockMind is built around exactly this split. With a Pro trial or subscription, you get a **personal AI investing agent**, a named analyst with its own workspace that handles the mechanical half while you keep the judgment half:
* **A [Morning Brief](/blog/crypto-morning-brief), by default around 8:00 in your timezone:** It explains what moved overnight, why it moved, and what matters for your holdings. The brief lands on your dashboard and by email.
* **A structured [research journey](/docs/capabilities/research-journey):** Explore scans for ideas, Analyze turns one into an analysis report, Verdicts gets second opinions from expert frameworks, and Track keeps watching saved ideas and holdings.
* **Monitoring and alerts:** Your agent watches holdings, tracked assets, and wider risks such as depegs and exploits. Daily [Fear & Greed](/blog/fear-greed-index-guide) and [Bitcoin dominance](/blog/bitcoin-dominance-explained) readings provide regime context. Where the optional Telegram companion is enabled, it supports chat and alert pushes; Morning Briefs still land on the dashboard and by email.
What it deliberately doesn't do is tell you what to buy or sell. Verdicts are research judgments, never trade instructions. BlockMind uses wallet and exchange connections only to read balances and positions. The agent can't trade, withdraw, or move funds even if you told it to. The judgment column of the table above stays yours by design.
The personal agent is included with Pro. See [current plans, introductory terms, and the card requirement](/docs/overview/plans-and-pricing) on the canonical pricing page.
If you are comparing ways to apply this split, read how a [dedicated crypto analyst compares with ChatGPT](/blog/ai-crypto-analyst-vs-chatgpt), what [AI portfolio monitoring](/blog/ai-portfolio-monitoring-crypto) changes after the first report, and how [BlockMind compares with CoinStats](/blog/blockmind-vs-coinstats). If the category itself is new, start with [what an AI crypto agent does all day](/blog/what-is-an-ai-crypto-agent).
## Frequently Asked Questions [#frequently-asked-questions]
### Which crypto research tasks should AI automate? [#which-crypto-research-tasks-should-ai-automate]
AI is best used for gathering data, cross-referencing sources, monitoring changes, and flagging known patterns. Those tasks are repetitive, measurable, and easy to audit against the underlying evidence.
### Which crypto research decisions should stay human? [#which-crypto-research-decisions-should-stay-human]
Keep founder quality, narrative timing, incentive design, thesis invalidation, and position sizing under human control. Those decisions depend on your goals, context, and willingness to bear the downside.
### Can AI replace a crypto research analyst? [#can-ai-replace-a-crypto-research-analyst]
AI can replace much of the repetitive gathering and monitoring, but it should not replace accountable judgment. Our guide to [whether AI can replace a crypto research analyst](/blog/can-ai-replace-crypto-research-analyst) explains the boundary in more detail.
### Can a BlockMind agent trade or move my funds? [#can-a-blockmind-agent-trade-or-move-my-funds]
No. BlockMind uses wallet and exchange connections only to read balances and positions. Your agent can research, monitor, and explain, but it cannot trade, withdraw, or move funds.
## The Bottom Line [#the-bottom-line]
"AI vs manual" is the wrong frame. The right frame is **AI for collection, humans for conviction**. Automate the gathering, cross-referencing, monitoring, and pattern flags: the work that is mechanical, repetitive, and punishing to do consistently by hand. Keep the founder reads, narrative timing, incentive analysis, and every sizing decision.
The manual purist does great research on two tokens and none on the other eight. The full automator outsources the one thing that was actually their edge. The investor who splits the work covers all ten and still makes every call themselves.
## Keep reading [#keep-reading]
- [Manual DYOR checklist: the 12 steps most people skip](https://docs.blockmind.app/blog/manual-dyor-checklist-12-steps)
- [How long does it take to research a crypto token?](https://docs.blockmind.app/blog/how-long-to-research-crypto-token)
- [The research journey: Explore → Analyze → Verdicts → Track](https://docs.blockmind.app/docs/capabilities/research-journey)
## Sources [#sources]
1. [NIST: AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework), accessed July 2026.
2. [CFTC: Customer Advisory on AI Trading Bots](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html), January 2024.
3. [SEC, NASAA, and FINRA: Artificial Intelligence and Investment Fraud](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/artificial-intelligence-fraud), January 2024.
---
# Best AI Crypto Research Tools in 2026: An Evidence-Based Comparison
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: commercial-investigation
Primary keyword: best AI crypto research tools
Keywords: best AI crypto research tools, AI crypto research platform, crypto AI tools, AI crypto analyst, crypto research software
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** There is no universally best AI crypto research tool. ChatGPT is the broadest general researcher; Messari Copilot and Kaito Pro are strong for curated crypto information retrieval; Nansen AI is built around on-chain intelligence and an execution surface; CoinStats AI Agent combines research with broad crypto portfolio tracking; and BlockMind is designed for a persistent, portfolio-aware analyst that briefs and monitors you. Choose the research job first, then verify the product against official documentation.
***
The best AI crypto research tool is the one that closes *your* evidence gap. A trader following wallet flows needs a different system from an analyst comparing protocol fundamentals. Someone with assets across hundreds of wallets and exchanges has a different constraint from someone who wants one agent to remember a thesis and revisit it next week.
This comparison was researched on **July 10, 2026**. It is based on public product pages, help centers, and pricing documentation. We did **not** create paid accounts, run standardized prompts, measure answer accuracy, or test support. Consequently, this is a documented-capability comparison, not a hands-on review. Vendor benchmark claims are identified as vendor claims rather than treated as independent proof.
For the underlying human-versus-automation decision, read [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research). For narrower product questions, see [AI Crypto Analyst vs ChatGPT](/blog/ai-crypto-analyst-vs-chatgpt) and the sourced [BlockMind vs CoinStats comparison](/blog/blockmind-vs-coinstats).
## The short list [#the-short-list]
| Tool | Best documented fit | Distinctive capability | Important limitation to verify |
| ------------------------- | ------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------ |
| **ChatGPT Deep Research** | Broad, multi-domain investigations | User-selected websites, files, apps, research plans, and cited reports | Crypto data is not portfolio-native unless supplied or connected; availability and limits vary by plan |
| **Messari Copilot** | Curated crypto research and market intelligence | Answers grounded in Messari research, market, fundraising, and on-chain datasets with citations | Deliberately crypto-specific; dataset and access depth vary by plan |
| **Kaito Pro** | Searching fragmented crypto information and social narratives | Crypto-native search, semantic retrieval, summarization, and AI Copilot | Public documentation is clearer on discovery than on personal portfolio accounting |
| **Nansen AI** | On-chain wallet, entity, and flow research | Nansen-labeled on-chain data, portfolio context, alerts, and an execution-oriented product surface | Execution changes the permission and operational-risk question; confirm exactly what you authorize |
| **CoinStats AI Agent** | Broad crypto tracking plus on-demand AI analysis | Portfolio-aware research across on-chain, social, technical, web, and exchange data | Vendor-published benchmark is not independent; suggestions are not a substitute for personal judgment |
| **BlockMind** | Persistent portfolio research, daily briefs, and ongoing monitoring | Named analyst, private workspace, Morning Brief, Notebook, expert verdicts, and recurring checks | Direct connection coverage is narrower than broad aggregators; no native mobile app; it never trades |
## Methodology [#methodology]
We evaluated only capabilities that a vendor documents publicly. The **FITS** framework asks four questions:
1. **Freshness:** What current sources can the tool reach, and does the vendor describe update timing?
2. **Investigation depth:** Can it conduct multi-step research, query structured data, compare evidence, or only summarize a prompt?
3. **Traceability:** Does it return citations or source links that a reader can audit?
4. **Situation awareness:** Can it use your portfolio, saved theses, alerts, or recurring context without rebuilding that context every time?
Execution is recorded separately because it is not automatically a benefit. A research system that can place orders has a different risk boundary from one that can only read data. BlockMind, for example, uses connected portfolio data for balance-only analysis and **cannot trade, withdraw, transfer, or sign wallet transactions**. See [Trust and security](/docs/security) for the exact boundary.
We did not assign a single numeric winner. Public feature lists do not establish answer accuracy, latency, uptime, or support quality. A score with false precision would hide that gap.
## Broad research: ChatGPT Deep Research [#broad-research-chatgpt-deep-research]
[OpenAI's current Deep Research documentation](https://help.openai.com/en/articles/10500283-deep-research-daq) says it can search the public web or sites you specify, use uploaded files and enabled apps, propose a research plan, and return a structured report with citations. That makes it the broadest option here: it can connect token mechanics to software architecture, macroeconomics, legal documents, or an industry's competitive landscape in one investigation.
Its strength is control over the research brief. You can require primary sources, limit the investigation to official domains, upload a protocol paper, and ask for counterevidence. It does not automatically know your wallets, position sizes, or prior theses unless you provide or connect that context.
OpenAI explicitly acknowledges that Deep Research can still hallucinate, make incorrect inferences, struggle to distinguish authoritative information from rumors, and calibrate confidence poorly. Citations make verification possible; they do not make verification optional. Check [current ChatGPT plans](https://openai.com/chatgpt/pricing/) instead of relying on a price quoted in an article.
**Choose it when:** the question crosses domains or you want tight control over sources.
**Do not choose it solely because:** a general model can produce polished prose. Fluency is not crypto data coverage.
## Messari-grounded answers: Messari Copilot [#messari-grounded-answers-messari-copilot]
[Messari's Copilot documentation](https://docs.messari.io/user-guides/messari-copilot/messari-copilot) describes an AI assistant that retrieves from Messari's curated research, market intelligence, fundraising data, on-chain metrics, and selected third-party content. It says answers include citations and that data can be roughly 15 minutes behind the latest events.
That is a useful fit when your main problem is navigating a large, curated crypto research terminal. The source pool is narrower than the open web by design, which can improve consistency but may omit a primary document or dissenting source outside the collection. A serious workflow should still open the cited governance proposal, filing, protocol documentation, or on-chain record.
Messari says Copilot is crypto-specific, which fits fundraising, protocol research, and sector mapping but constrains cross-domain questions. Review [Messari's current plan page](https://messari.io/pricing) for access and limits.
**Choose it when:** you already value Messari's research and datasets and want a conversational retrieval layer over them.
## Best for crypto-native information discovery: Kaito Pro [#best-for-crypto-native-information-discovery-kaito-pro]
[Kaito's product documentation](https://kaito-ai.gitbook.io/product-docs) positions the platform as a Web3 information layer that indexes fragmented sources and adds semantic search, summarization, analytics, and an AI Copilot. Its documented use cases emphasize finding relevant research and social content without manually searching many channels.
That makes Kaito especially relevant to narrative research: who is discussing a topic, which themes are accelerating, and what source material sits behind a project or ticker. Social attention is evidence of attention—not evidence that a claim is true or that a token is valuable. Use discovery output to locate sources, then verify material claims against protocol documentation and on-chain data.
The public documentation we reviewed is stronger on search and market intelligence than on personal position accounting. If portfolio-aware analysis is essential, confirm the exact supported workflow before subscribing. See [Kaito's product and access information](https://www.kaito.ai/) for current availability.
**Choose it when:** fragmented crypto information and narrative discovery are the bottleneck.
## Best for labeled on-chain intelligence: Nansen AI [#best-for-labeled-on-chain-intelligence-nansen-ai]
[Nansen AI](https://nansen.ai/) is built around on-chain data, labeled wallets and entities, portfolio information, research, and an execution-oriented interface. Nansen's [Smart Alerts documentation](https://academy.nansen.ai/en/help/articles/6239622-ai-smart-alerts-101) also documents alerts for addresses, token flows, exchange flows, and contract interactions.
This is the clearest fit when your research question begins with *who moved what on-chain?* Labels can make public blockchain activity more legible. They remain analytical classifications: a transfer into an exchange-associated address does not prove an imminent sale, and a labeled profitable wallet does not make its next action wise.
Nansen's execution capability is a material product difference, not a line-item bonus. Before enabling any action surface, inspect wallet approvals, signing steps, custody model, supported networks, fees, and the exact boundary between recommendation and execution. The safest research connection is the least-privileged one that satisfies the task. Consult [Nansen's current Pro plan documentation](https://academy.nansen.ai/articles/9412804-about-nansen-pro) directly.
**Choose it when:** labeled wallet behavior and on-chain flows are central to the decision.
## Best broad tracker with an AI copilot: CoinStats AI Agent [#best-broad-tracker-with-an-ai-copilot-coinstats-ai-agent]
CoinStats launched its current [AI Agent](https://coinstats.app/blog/ai/) in April 2026. Its official overview says specialized agents can search news, social data, blockchain activity, exchange data, technical indicators, and the user's connected portfolio. It also documents portfolio analysis, wallet analysis, risk scanning, and backtesting.
CoinStats' larger product is a broad crypto portfolio manager, so this option makes sense when aggregation and mobile access matter alongside research. The company publishes a benchmark comparing its agent with general-purpose systems, but the benchmark was designed and published by CoinStats. Its open methodology is useful; independent replication would be stronger evidence. We do not repeat its headline result as an established product ranking.
The product page says the agent can suggest portfolio adjustments. Treat those as research output, not personalized fiduciary advice. Check assumptions, costs, tax effects, liquidity, and your own constraints. Review [CoinStats plans](https://coinstats.app/pricing/) for current access rather than relying on a static price here.
**Choose it when:** you want extensive crypto portfolio aggregation, native apps, and an on-demand AI layer in one product.
## Best for persistent, non-trading portfolio research: BlockMind [#best-for-persistent-non-trading-portfolio-research-blockmind]
BlockMind gives Pro members a named personal analyst with a private workspace. The agent can research crypto, stocks, and commodities; use connected holdings as context; retain durable notes in a [Notebook](/docs/memory/notebook); produce a portfolio-aware [Morning Brief](/docs/automations/morning-brief); and run scheduled [monitoring and alerts](/docs/automations/monitoring-and-alerts).
The product's research journey is explicit: **Explore → Analyze → Verdicts → Track**. That is useful when the missing ingredient is continuity rather than another one-off answer. A saved idea can become a structured analysis, face several expert frameworks, and remain monitored after the report.
There are real tradeoffs. BlockMind's direct wallet and exchange coverage is smaller than broad aggregators such as CoinStats, and the product is a responsive web app rather than a native iOS or Android app. Monitoring runs periodically, not at tick-level speed. Most importantly, the agent is intentionally not an execution system: it cannot trade or move funds and never tells you what to buy or sell.
See [what your agent can do](/docs/overview/what-your-agent-can-do) and [plans and pricing](/docs/overview/plans-and-pricing) for the current product contract.
**Choose it when:** you want research, memory, briefs, and monitoring tied together under a read-only boundary.
## A practical decision table [#a-practical-decision-table]
| Your main job | Start with | Why | Verification question |
| --------------------------------------------------------------- | --------------------- | ----------------------------------------------------------- | ------------------------------------------------------------------------------------------ |
| Research a protocol across technical, legal, and market sources | ChatGPT Deep Research | Broad source control and cross-domain synthesis | Did the report cite the deployed docs, filings, and governance records? |
| Query curated crypto market and fundraising data | Messari Copilot | Conversational access to Messari datasets | Is the underlying dataset included in your plan? |
| Map a narrative and discover relevant crypto content | Kaito Pro | Crypto-native search and semantic discovery | Can you trace each important claim to a primary source? |
| Investigate wallet flows and labeled entities | Nansen AI | On-chain labels and alerting | What does the label mean, and what alternative explanation fits the transfer? |
| Track many crypto accounts and ask portfolio questions | CoinStats AI Agent | Aggregation plus AI and native apps | Are all venues, transaction types, and cost bases syncing correctly? |
| Maintain a recurring research habit around your holdings | BlockMind | Persistent agent, Notebook, briefs, and periodic monitoring | Are your connected holdings complete, and is periodic—not real-time—monitoring sufficient? |
## The FITS worksheet: an auditable way to choose [#the-fits-worksheet-an-auditable-way-to-choose]
Score only what you can demonstrate in a trial or official documentation. Use 0 for absent, 1 for partial, and 2 for clearly supported.
| Test | Weight | Evidence to save |
| ---------------------------------------- | -----: | ------------------------------------------- |
| Retrieves current primary sources | 3 | Links from a dated sample report |
| Cites claims at paragraph level | 3 | Screenshot or exported answer |
| Covers the structured data you need | 3 | Supported-data documentation |
| Understands your portfolio accurately | 3 | Reconciled holdings against source accounts |
| Preserves theses and prior decisions | 2 | Memory or notebook documentation |
| Supports useful alerts or recurring work | 2 | Alert types, cadence, and delivery docs |
| Exposes clear permission boundaries | 3 | Security and connection documentation |
| Makes export or audit easy | 1 | Report and citation export workflow |
Calculate a weighted coverage percentage as: `sum of score × weight ÷ sum of maximum score × weight × 100`.
**Hypothetical example:** A tool earns 29 weighted points out of a possible 40. Its coverage is `29 ÷ 40 × 100 = 72.5%`. That number is not an accuracy score. It only says the product demonstrated 72.5% of *your documented requirements*. A missing permission boundary can still be disqualifying even if the total is high.
## Limitations and counterevidence [#limitations-and-counterevidence]
* We assessed documentation, not real-world answer quality. Product pages are written by vendors.
* Features, plan access, rate limits, and prices can change after July 10, 2026.
* A citation can be stale, irrelevant, or misread. Open it.
* Portfolio sync can omit unsupported chains, derivatives, staking rewards, internal transfers, or historical transactions.
* AI systems can hallucinate, misidentify similarly named tokens, and infer causation from coincident events.
* On-chain transparency does not reveal every owner's identity or intent.
* More automation is not always better. Execution access increases the consequences of a wrong output or compromised credential.
## The Bottom Line [#the-bottom-line]
Start with the research job, not the brand. Choose ChatGPT for cross-domain investigations, Messari for curated crypto intelligence, Kaito for information discovery, Nansen for labeled on-chain analysis, CoinStats for broad tracking plus AI, or BlockMind for a persistent analyst that briefs and monitors under a non-trading boundary.
Then run the same three questions through your shortlist, save the sources, reconcile any portfolio data, and grade the result with FITS. The best AI crypto research tool is not the one that sounds most confident. It is the one whose evidence you can inspect and whose permissions match the job.
This article is for research and education, not financial advice. No AI system can determine what you should buy or sell.
## Sources [#sources]
1. [OpenAI: Deep Research in ChatGPT](https://help.openai.com/en/articles/10500283-deep-research-daq)
2. [OpenAI: Deep Research limitations and updates](https://openai.com/index/introducing-deep-research/)
3. [Messari: Copilot documentation](https://docs.messari.io/user-guides/messari-copilot/messari-copilot)
4. [Kaito: Product documentation](https://kaito-ai.gitbook.io/product-docs)
5. [Nansen: AI product overview](https://nansen.ai/)
6. [Nansen: AI Smart Alerts documentation](https://academy.nansen.ai/en/help/articles/6239622-ai-smart-alerts-101)
7. [CoinStats: AI Agent capability overview](https://coinstats.app/blog/ai/)
8. [BlockMind: What your agent can do](/docs/overview/what-your-agent-can-do)
9. [BlockMind: Trust and security](/docs/security)
---
# Best Crypto Portfolio Trackers in 2026: Six Workflows Compared
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: commercial-investigation
Primary keyword: best crypto portfolio trackers
Keywords: best crypto portfolio trackers, crypto portfolio tracker comparison, best crypto tracking app, DeFi portfolio tracker, crypto portfolio monitoring
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** CoinStats is the strongest documented fit for broad crypto aggregation plus native apps; Delta for a cross-asset mobile overview; CoinTracker for crypto tax records tied to portfolio tracking; Zerion for wallet-native DeFi positions; CoinGecko for lightweight manual portfolios and price alerts; and BlockMind for portfolio-aware research, Morning Briefs, and periodic monitoring. The best tracker is the one that reconciles *your* accounts and supports *your* workflow.
***
No portfolio tracker supports every chain, exchange product, DeFi position, derivative, cost-basis method, and alert. Breadth claims also do not establish correctness. A tracker can connect to an account and still omit staking rewards, internal transfers, delisted assets, derivatives, or old transactions because the source API does not expose them.
This comparison was checked on **July 10, 2026**. It relies on official product pages and help centers. We did not open paid accounts, connect live portfolios, test sync accuracy, measure support, or run a standard mobile trial. BlockMind publishes this article and is included as a first-party product; readers should apply extra scrutiny to our own characterization and verify it against the linked documentation.
For AI research products beyond portfolio aggregation, use the cluster pillar, [Best AI Crypto Research Tools](/blog/best-ai-crypto-research-tools). This page owns the narrower tracker and portfolio-workflow comparison.
## Methodology [#methodology]
We used the **TRACK** framework and credited only publicly documented capabilities:
1. **T — Total coverage:** exchanges, wallets, chains, DeFi, NFTs, traditional assets, CSV, and manual entry.
2. **R — Reconciliation:** transaction history, balances, cost basis, P\&L, spam handling, and documented sync gaps.
3. **A — Analysis:** allocation, performance, tax, risk, AI research, and look-through context.
4. **C — Cadence:** sync frequency, price and event alerts, briefs, recurring monitoring, and mobile delivery.
5. **K — Key boundary:** public addresses, OAuth, API-key scope, custody, trading, and withdrawal authority.
We do not assign an overall numeric score. Official feature lists cannot prove sync accuracy, uptime, latency, privacy practice in operation, or how well a particular account imports. Pricing and plan limits change, so this article links to official pricing pages instead of embedding volatile prices.
## At-a-glance comparison [#at-a-glance-comparison]
| Tracker | Best documented fit | Coverage shape | Analysis and alerts | Key limitation to verify |
| ------------------ | --------------------------------------------------------- | ----------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
| **BlockMind** | Portfolio-aware analyst, briefs, and ongoing research | 9 exchanges, wallets on 8 networks, CSV, manual; crypto plus cross-asset research | Agent analysis, Morning Brief, periodic monitoring, in-app/email/Telegram alerts | Narrower direct integrations than broad aggregators; no native app or tax reporting; monitoring is periodic |
| **CoinStats** | Broad crypto aggregation with native apps and AI | Official site advertises 300+ wallets/exchanges across 120+ blockchains; DeFi and NFTs | P\&L, analytics, alerts, portfolio-aware AI Agent, iOS/Android | Reconcile unsupported transaction types and source-API gaps; AI suggestions need human verification |
| **Delta by eToro** | Mobile overview across crypto and traditional investments | Crypto wallets/exchanges plus stocks, funds, forex, indices, and documented broker/bank connections | Portfolio diversity, performance, event/price alerts, mobile-first insights | Many connections sync periodically; official docs describe history gaps and synthetic balance adjustments |
| **CoinTracker** | Crypto tax reporting plus year-round portfolio tracking | Official site advertises 500+ exchanges, wallets, and blockchains plus CSV and DeFi coverage | Tax reports, cost basis, gains/losses, allocation and performance | Plans depend on transaction volume and tax needs; complex transactions still require reconciliation |
| **Zerion** | On-chain wallet and DeFi portfolio | Ethereum, Solana, major EVM networks, DeFi positions, NFTs; public-address tracking | DeFi P\&L, activity notifications, wallet following; mobile and browser wallet | Does not aggregate centralized-exchange balances in the same way; wallet includes swap/signing capabilities |
| **CoinGecko** | Lightweight manual tracking and market alerts | User-created portfolios across listed coins; manual holdings rather than exchange-account aggregation | Prices, P\&L inputs, news, categories, one-time/recurring alerts, mobile widgets | Manual transaction maintenance; not a full tax ledger or connected DeFi/account reconciler |
## Best for a persistent portfolio-aware analyst: BlockMind [#best-for-a-persistent-portfolio-aware-analyst-blockmind]
BlockMind combines tracking with a named personal analyst. Portfolios can come from public wallets, provider-scoped read-only exchange connections where supported, CSV import, or manual entry. Current direct support covers nine exchanges and wallet tracking across eight networks, documented in [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
The differentiator is what happens after aggregation. The agent can answer allocation, exposure, performance, and position-change questions; use holdings in research; prepare a personalized Morning Brief; and run monitoring passes every few hours. It can save durable notes and theses in a Notebook.
The tradeoffs are clear. CoinStats and CoinTracker document much broader direct aggregation. Delta has native mobile apps and cross-asset account connections. BlockMind does not provide tax reports, transaction export, NFT tracking, or trade execution. Monitoring is not tick-level. It is best when the desired outcome is an ongoing research relationship rather than the largest integration count.
BlockMind uses connections only to read balances and positions. The agent cannot trade, withdraw, transfer, or sign transactions. Firi does not offer a provider-scoped read-only key; the application enforces balance-only behavior for that credential. Review [Trust and security](/docs/security) and [current plans](/docs/overview/plans-and-pricing).
**Choose it when:** the central job is “explain and monitor what I own,” not tax filing or maximum account breadth.
## Best for broad crypto aggregation and native apps: CoinStats [#best-for-broad-crypto-aggregation-and-native-apps-coinstats]
CoinStats' official [integrations page](https://coinstats.app/integrations/) advertises more than 300 wallet and exchange connections across more than 120 blockchains. Its product includes centralized-exchange, wallet, DeFi, and NFT tracking, plus P\&L and portfolio analytics on web, iOS, and Android.
CoinStats also launched a portfolio-aware [AI Agent](https://coinstats.app/blog/ai/) in April 2026, documenting news, on-chain, social, technical, exchange, portfolio-analysis, and backtesting capabilities. That makes it broader than a passive tracker.
Breadth still needs reconciliation. Confirm that the exact exchange products, chains, protocols, derivatives, rewards, and historical periods you use are imported. AI-generated portfolio suggestions are not a substitute for personalized professional advice. Check [current CoinStats plans](https://coinstats.app/pricing/) and transaction/portfolio limits directly.
**Choose it when:** broad crypto account coverage, DeFi/NFT visibility, native mobile access, and an AI copilot matter together.
## Best cross-asset mobile overview: Delta by eToro [#best-cross-asset-mobile-overview-delta-by-etoro]
Delta tracks cryptocurrencies alongside stocks, funds, forex, indices, and NFTs. Its [official product description](https://support.delta.app/en/articles/1435355-what-is-delta-by-etoro) positions the app as a consolidated view across asset types, and its current [features page](https://delta.app/en/features) documents account sync, performance, diversity analysis, market-move explanations, and alerts.
The breadth runs beyond crypto: Delta's broker-connection documentation says it can link to more than 1,600 brokers and banks. The same documentation is unusually candid about limitations. Some providers expose only partial history, some require reauthentication, and most broker connections sync every 24 hours. Delta may create automatic balance-adjustment transactions when imported history does not match the current balance, which can distort cost basis until reconciled.
Delta says it is a read-only tracker and does not hold or move funds. For exchange and wallet connection limits and plan access, consult the [Delta help center](https://support.delta.app/en/collections/860856-delta-pro) and in-app pricing, which can vary by app store and region.
**Choose it when:** a unified mobile view across crypto and traditional investments is more important than crypto tax depth or an always-on analyst.
## Best for crypto tax records and portfolio tracking: CoinTracker [#best-for-crypto-tax-records-and-portfolio-tracking-cointracker]
CoinTracker combines portfolio tracking with tax reporting. Its [features page](https://www.cointracker.io/features) advertises support across more than 500 exchanges, wallets, and blockchains, along with smart contracts, dapps, and CSV imports. Its help center documents current value, asset-level holdings, allocation, performance, gains/losses, and tax-report features.
This is the clearest fit when the portfolio record must feed a tax workflow. Transaction count, tax years, tax-loss tools, and support level vary by plan; check [current CoinTracker plan features](https://support.cointracker.io/hc/en-us/articles/23466080743825-CoinTracker-Plan-Features) and pricing.
Tax software is only as accurate as the ledger. DeFi, staking, wrapped assets, bridge transfers, spam, missing history, and owned-wallet transfers can require classification or correction. CoinTracker's [security page](https://www.cointracker.io/security) says Coinbase uses read-only OAuth and other exchange connections use read-access API keys, with no request for wallet private keys.
**Choose it when:** tax reporting and a reconciled transaction ledger are the main jobs.
## Best for wallet-native DeFi tracking: Zerion [#best-for-wallet-native-defi-tracking-zerion]
Zerion is both a self-custody wallet and an on-chain portfolio tracker. Its [official product page](https://www.zerion.io/) documents tokens, DeFi positions, debts, rewards, NFTs, and transaction history across Ethereum, Solana, and major EVM networks. The web app can also watch public addresses without installing a wallet.
That makes Zerion useful for someone whose portfolio lives mainly on-chain. Protocol-level presentation can show positions that a simple token-balance tracker misses. It does not aggregate centralized-exchange balances in the same way; Zerion's own FAQ says the DeFi tracker works for non-custodial wallets and cannot show assets held on centralized exchanges.
The key boundary differs from a read-only dashboard because Zerion Wallet can also sign, swap, and bridge. Watching a public address is observation; connecting a self-custody wallet creates an interaction surface. Review every transaction before signing. Check [Zerion Premium](https://www.zerion.io/premium) for current paid features.
**Choose it when:** DeFi positions and wallet activity are the portfolio, and centralized-account aggregation is secondary.
## Best lightweight manual portfolio: CoinGecko [#best-lightweight-manual-portfolio-coingecko]
CoinGecko lets users create portfolios, enter holdings, group coins, view price performance and news, and set one-time or recurring alerts. Its [portfolio and price-alert guide](https://www.coingecko.com/learn/coingecko-crypto-price-alerts-portfolio) documents the workflow on web and mobile.
The advantage is low connection complexity. A manual portfolio can be useful when you want a watchlist-like valuation view without sharing exchange credentials. The cost is maintenance: buys, sells, transfers, cost basis, staking, and DeFi positions do not automatically reconcile like a connected ledger.
CoinGecko is a market-data and portfolio view, not a custody account. Verify the exact token contract and price source, particularly for long-tail assets. Check its [mobile and alert features](https://www.coingecko.com/en/mobile.html) for current availability.
**Choose it when:** you want a simple market watch and manually maintained holdings without account integrations.
## The reconciliation trial: how to test a tracker honestly [#the-reconciliation-trial-how-to-test-a-tracker-honestly]
Run the same bounded trial with each finalist. Do not judge by a demo portfolio.
1. Choose one exchange, one wallet, and one DeFi position you understand.
2. Record source balances at one timestamp.
3. Import through the least-privileged supported method.
4. Compare total value, quantities, liabilities, and missing assets.
5. Test five transaction types: spot trade, deposit, withdrawal, internal transfer, and staking/reward if applicable.
6. Check token identity by contract, not symbol.
7. Inspect spam filtering and low-liquidity valuation.
8. Compare cost basis and realized/unrealized P\&L definitions.
9. Wait through one documented sync cycle and repeat.
10. Disconnect and verify credential revocation and data-deletion behavior.
### Calculate reconciliation coverage [#calculate-reconciliation-coverage]
`coverage = correctly reconciled source value ÷ independently verified source value × 100`.
**Hypothetical:** Source accounts total $50,000. The tracker correctly represents $46,500, omits $2,000 of staked assets, and misprices a $1,500 long-tail token. Coverage is `46,500 ÷ 50,000 × 100 = 93%`.
The percentage does not measure tax accuracy, privacy, or insight quality. It tells you 7% of the verified value is unresolved in this trial. Save the missing categories; they matter more than the headline score.
## Workflow decision table [#workflow-decision-table]
| Your primary need | Start with | Also consider |
| ---------------------------------------------- | ----------- | -------------------------------------------------- |
| Broadest documented crypto account aggregation | CoinStats | CoinTracker if tax matters |
| Cross-asset mobile overview | Delta | BlockMind for cross-asset research context |
| Tax reports and transaction ledger | CoinTracker | Accountant review for complex activity |
| DeFi and public-wallet positions | Zerion | Explorer and protocol UIs for verification |
| Manual portfolio and price alerts | CoinGecko | Spreadsheet for custom accounting |
| Portfolio-aware briefs and research monitoring | BlockMind | Broader tracker if direct coverage is insufficient |
Using two products can be rational: one as the ledger of record and another for research. Avoid connecting the same high-permission credential to both. A tracker should use provider-scoped read-only access; see [Read-Only Crypto API Keys](/blog/read-only-api-keys-crypto-security).
## Limitations and counterevidence [#limitations-and-counterevidence]
* All capability and integration figures are vendor-published and can change after July 10, 2026.
* “Supported” may mean balances but not complete history, cost basis, derivatives, or every protocol.
* We did not test sync accuracy, security controls in operation, support, or app usability.
* Regional and app-store availability can differ.
* A read-only connection still exposes sensitive financial data.
* Portfolio valuations can be wrong when liquidity is thin or token identity is mis-mapped.
* Tax calculations require jurisdiction-specific treatment and may need professional review.
* AI analysis can hallucinate or use stale information; open the sources and check the math.
After choosing, run the [crypto portfolio health check](/blog/crypto-portfolio-health-check) before trusting concentration or performance output.
## The Bottom Line [#the-bottom-line]
There is no best crypto portfolio tracker independent of workflow. CoinStats wins documented crypto breadth, Delta cross-asset mobile overview, CoinTracker tax workflow, Zerion DeFi visibility, CoinGecko low-friction manual tracking, and BlockMind proactive portfolio research.
Shortlist by TRACK, then connect a small representative sample and measure reconciliation. The tracker that imports the most accounts is not necessarily the tracker that represents your actual exposure correctly.
This article is for research and education, not financial, tax, legal, or security advice. BlockMind never trades or moves funds.
## Sources [#sources]
1. [CoinStats: Supported integrations](https://coinstats.app/integrations/)
2. [CoinStats: AI Agent capabilities](https://coinstats.app/blog/ai/)
3. [Delta: Product scope](https://support.delta.app/en/articles/1435355-what-is-delta-by-etoro)
4. [Delta: Broker connection behavior and limitations](https://support.delta.app/en/articles/5765798-broker-account-connections)
5. [Delta: Exchange-connection limitations](https://support.delta.app/en/articles/1587431-limitations-to-exchange-account-connections)
6. [CoinTracker: Portfolio and tax features](https://www.cointracker.io/features)
7. [CoinTracker: Security and read-only connections](https://www.cointracker.io/security)
8. [Zerion: Wallet and DeFi portfolio tracking](https://www.zerion.io/)
9. [CoinGecko: Portfolio and alert workflow](https://www.coingecko.com/learn/coingecko-crypto-price-alerts-portfolio)
10. [BlockMind: Portfolios](/docs/data/portfolios)
11. [BlockMind: Monitoring and alerts](/docs/automations/monitoring-and-alerts)
---
# Bitcoin Dominance Explained: What It Tells You About the Market
Published: 2026-02-07
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: bitcoin dominance explained
Keywords: bitcoin dominance, BTC dominance, altcoin market
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Bitcoin dominance measures Bitcoin's share of total crypto market cap. High and rising dominance often means capital is favoring Bitcoin; low and falling dominance can signal stronger risk appetite for altcoins. Use it alongside the [Fear & Greed Index](/blog/fear-greed-index-guide) to understand market conditions instead of treating either reading as a trading signal.
***
Bitcoin dominance is Bitcoin's market cap as a percentage of the total crypto market. When BTC dominance is 60%, it means Bitcoin represents 60% of all crypto value. This metric tells you whether money is flowing toward Bitcoin or spreading into altcoins.
## How It's Calculated [#how-its-calculated]
The formula is simple:
```
BTC Dominance = (Bitcoin Market Cap / Total Crypto Market Cap) × 100
```
**Example:** If Bitcoin is worth $1.2 trillion and the total crypto market is $2 trillion, BTC dominance is 60%.
## What the Numbers Mean [#what-the-numbers-mean]
| Dominance | What It Signals |
| --------- | -------------------------------------------------- |
| Above 60% | Investors strongly favor Bitcoin over altcoins |
| 50–60% | Bitcoin leads, but market participation is broader |
| 40–50% | Altcoin interest is taking more market share |
| Below 40% | Altcoin participation is unusually high |
## Historical Patterns [#historical-patterns]
Bitcoin dominance has followed recognizable patterns through market cycles:
### 2017: The ICO Boom [#2017-the-ico-boom]
* Started the year near 90%
* Fell to roughly 38% during the ICO craze
* Thousands of new tokens diluted Bitcoin's share
* This extreme low preceded the 2018 crash
### 2018–2019: Bear Market [#20182019-bear-market]
* Rose back above 70%
* Investors fled to Bitcoin as altcoins collapsed
* "Flight to quality" meant BTC was treated as the safest major crypto asset
### 2021: Alt Season [#2021-alt-season]
* Dropped to around 40% during NFT and DeFi mania
* Ethereum and Solana gained significant market share
* Marked another local top
### 2022 to Mid-2026: Recovery and Consolidation [#2022-to-mid-2026-recovery-and-consolidation]
* Dominance climbed from the post-FTX lows as institutional adoption and the "digital gold" narrative strengthened
* [CoinGecko recorded BTC dominance near 56% on July 10, 2026](https://www.coingecko.com/en/charts/bitcoin-dominance), compared with roughly 61% one year earlier
* Current readings move with the market; BlockMind's Market Indicators page refreshes its value once daily
## What Rising BTC Dominance Means [#what-rising-btc-dominance-means]
When dominance is increasing:
1. **Investors may be de-risking:** Capital is moving from volatile altcoins to Bitcoin.
2. **Altcoins are underperforming:** BTC is outpacing the rest of the market.
3. **Market participation is narrowing:** More value is concentrated in the largest asset.
4. **A rotation may be underway:** Capital can move to BTC before risk appetite returns.
**Strategy consideration:** During sustained rises in dominance, Bitcoin has often outperformed altcoins. That relationship is historical context, not a guarantee.
## What Falling BTC Dominance Means [#what-falling-btc-dominance-means]
When dominance is decreasing:
1. **Risk appetite may be rising:** Investors are hunting for higher returns.
2. **Altcoins are outperforming:** Smaller assets are gaining market share.
3. **Market participation is broadening:** More capital is moving beyond Bitcoin.
4. **Speculation may be increasing:** Late-stage euphoria can appear before corrections.
**Strategy consideration:** During falling dominance, some altcoins may outperform Bitcoin. Be cautious because extreme lows have appeared near overheated markets. This is also when weak projects can attract attention, so review the [signs of a rug pull](/blog/5-signs-crypto-rug-pull) before taking risk.
## How to Use BTC Dominance in Your Strategy [#how-to-use-btc-dominance-in-your-strategy]
### 1. Confirm Market Phase [#1-confirm-market-phase]
High and rising dominance often indicates early or late cycle phases. Low and falling dominance suggests mid-cycle euphoria.
### 2. Time Your Rotations [#2-time-your-rotations]
Some investors use the dominance trend as one input when assessing relative performance:
* Dominance rising → Bitcoin is gaining market share
* Dominance falling → Altcoins are gaining market share
Do not rotate solely because this one metric moved. Check price direction, liquidity, sentiment, and your own risk limits too.
### 3. Set Expectations [#3-set-expectations]
During Bitcoin seasons, don't expect altcoins to outperform. During alt seasons, understand the increased risk.
### 4. Combine with Sentiment [#4-combine-with-sentiment]
BTC dominance is most powerful when paired with the [Fear & Greed Index](/blog/fear-greed-index-guide). For example:
* **High dominance + extreme fear** = Defensive sentiment with Bitcoin relatively favored
* **Low dominance + extreme greed** = Overheated conditions and a possible top
* **Falling dominance + rising greed** = Alt season in full swing, but be cautious
### 5. Spot Extremes [#5-spot-extremes]
Extreme readings above 70% or below 40% have appeared near important turning points. They are not guarantees, but they are worth investigating.
## Limitations to Understand [#limitations-to-understand]
BTC dominance isn't a crystal ball. Keep in mind:
* **Stablecoins distort the metric:** [DefiLlama put their total market cap near $312 billion in July 2026](https://defillama.com/stablecoins), so they materially affect the denominator.
* **It's descriptive, not predictive:** It tells you what's happening, not what will happen.
* **Timeframes matter:** Daily noise and monthly trends tell different stories.
* **It doesn't capture everything:** DeFi TVL, on-chain activity, and other metrics add context.
## Tracking BTC Dominance with BlockMind [#tracking-btc-dominance-with-blockmind]
BlockMind refreshes BTC dominance once daily alongside the [Fear & Greed Index](/blog/fear-greed-index-guide), the Altcoin Season Index, and its Contrarian Leverage Index.
Create a free BlockMind account and open the [Market Indicators page](https://blockmind.app/market/indicators) to see BTC dominance, Fear & Greed, Altcoin Season, and BlockMind's in-house Contrarian Leverage Index with their history.
With a Pro trial or subscription, your personal AI investing agent can do the checking for you. Its dashboard keeps the indicators in one carousel, and you can ask in chat what today's combination means for your holdings. Your agent scans the market daily and brings relevant changes into a [Morning Brief](/blog/crypto-morning-brief) delivered to your dashboard and email.
Learn more about [how BlockMind's indicators work](/docs/classic/market-indicators). These readings are research context, not financial advice. Neither the indicators nor your agent will tell you what to buy or sell.
## Frequently Asked Questions [#frequently-asked-questions]
### What is a good Bitcoin dominance level? [#what-is-a-good-bitcoin-dominance-level]
There's no universally "good" level because the meaning depends on the broader market and your strategy. Higher dominance confirms Bitcoin's relative strength, while falling dominance shows altcoins gaining share. Neither is an automatic opportunity.
### Does low Bitcoin dominance mean altcoins will go up? [#does-low-bitcoin-dominance-mean-altcoins-will-go-up]
Not necessarily. Low BTC dominance means altcoins are gaining market share *relative* to Bitcoin, but during bear markets, both Bitcoin and altcoins can fall while dominance stays flat. Check overall market direction alongside dominance.
### How often does Bitcoin dominance change? [#how-often-does-bitcoin-dominance-change]
BTC dominance changes continuously as crypto prices fluctuate. Major shifts (10%+ swings) typically play out over months, not days. For strategy decisions, focus on weekly and monthly trends rather than daily readings.
## The Bottom Line [#the-bottom-line]
Bitcoin dominance is one of the most useful macro indicators in crypto. It won't tell you what to buy, but it will help you understand the market environment:
* **High dominance** = Defensive, Bitcoin-favoring conditions
* **Low dominance** = Aggressive, altcoin-favoring conditions
* **Extreme readings** = Conditions worth investigating for possible turning points
Use it as one input alongside your own research. No single metric tells the whole story.
For a complete top-down process, put dominance inside the [crypto market analysis framework](/blog/crypto-market-analysis-guide), then compare it with [on-chain evidence](/blog/what-is-on-chain-analysis-crypto) and sentiment rather than reading it alone.
## Sources [#sources]
1. [CoinGecko: Bitcoin Dominance chart](https://www.coingecko.com/en/charts/bitcoin-dominance), accessed July 2026.
2. [CoinMarketCap: Bitcoin Dominance methodology](https://support.coinmarketcap.com/hc/en-us/articles/360043395912-Bitcoin-Dominance-Percentage), accessed July 2026.
3. [DefiLlama: Stablecoins](https://defillama.com/stablecoins), accessed July 2026.
---
# BlockMind vs CoinStats (2026): Which AI Crypto Analyst Fits You?
Published: 2026-03-09
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: comparison
Primary keyword: BlockMind vs CoinStats
Keywords: BlockMind vs CoinStats, CoinStats alternative, AI crypto portfolio tracker
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** CoinStats is the stronger choice when you want broad crypto tracking, native mobile apps, and an AI research copilot inside a mature portfolio manager. BlockMind is built for a different job: a dedicated personal investing agent that keeps working between visits, sends a Morning Brief, and turns research into explicit expert verdicts. Both now offer portfolio-aware AI. The useful comparison is no longer "tracker versus analyst," but what kind of analyst you want.
***
CoinStats and BlockMind overlap more than they used to. Both connect read-only portfolio data, both answer questions about your holdings, and both offer AI-assisted crypto research. CoinStats introduced its [AI Agent](https://coinstats.app/blog/ai/) in April 2026, so any comparison that treats it as a charting app with only price predictions is out of date.
The products still make different tradeoffs. CoinStats starts with a comprehensive crypto tracker and adds an on-demand research copilot. BlockMind starts with a dedicated agent and gives it a persistent workspace, scheduled monitoring, a daily brief, and structured expert review.
## Methodology [#methodology]
This comparison was reviewed on July 10, 2026 using the companies' current public product, integration, help, and pricing pages. It compares documented workflow fit, coverage, permissions, delivery surfaces, limits, and advertised US pricing. It is not a hands-on performance benchmark, and it does not infer unlisted features. Where CoinStats' own pages disagree, both values are shown instead of choosing the more favorable one. Regional prices, promotions, app-store billing, and shipped features can change; confirm current terms before purchase.
To compare the broader categories before narrowing to these two products, use the [AI crypto research tools methodology](/blog/best-ai-crypto-research-tools).
## BlockMind vs CoinStats at a glance [#blockmind-vs-coinstats-at-a-glance]
| Question | BlockMind | CoinStats |
| ------------------------------ | ------------------------------------------------------------------------- | ---------------------------------------------------------------------------------- |
| Best for | A proactive personal investing agent | Broad crypto tracking with an AI copilot |
| AI access | Pro | Premium and Degen |
| Portfolio awareness | Yes | Yes |
| Proactive daily brief | Personalized Morning Brief in the dashboard and by email | Not listed in the current AI Agent or plan documentation |
| Research style | Explore, Analyze, Verdicts, Track | Prompt-based multi-agent research, portfolio analysis, and backtesting |
| Asset scope | Crypto, stocks, and commodities | Crypto, DeFi, and NFTs |
| Connected coverage | 9 exchange API integrations and wallets across 8 networks | 300+ wallets and exchanges across 120+ blockchains |
| Mobile access | Responsive web app | Web, iOS, and Android |
| Entry price for the AI product | [See current BlockMind plans and terms](/docs/overview/plans-and-pricing) | 7-day free trial on the yearly plan, then $13.99/month billed annually for Premium |
| Free tracking | Up to 5 combined portfolios and watchlists | 10 portfolios and 20,000 transactions |
Pricing and published limits were checked on July 10, 2026. App-store billing, promotions, and regional prices can differ.
## The real difference: copilot or dedicated agent [#the-real-difference-copilot-or-dedicated-agent]
CoinStats describes its AI Agent as a crypto research copilot. When you ask a question, specialized agents search news, scan social data, review blockchain and exchange data, inspect your portfolio, and synthesize an answer. It can explain market moves, analyze wallets, review sentiment, answer portfolio questions, backtest ideas, and run code. That is considerably more capable than the price-prediction feature CoinStats offered before the agent launched.
BlockMind is designed around continuity. You name one personal AI investing agent, and it works from its own workspace. It knows your portfolio, can research assets on demand, wakes on a schedule to monitor what matters, and sends a personalized Morning Brief every day. Its research journey is explicit:
1. **Explore** scans markets for ideas.
2. **Analyze** turns an idea into a structured report.
3. **Verdicts** asks a panel of experts for a call: approved, watch, wait, or reject.
4. **Track** keeps watching holdings and followed verdicts.
That distinction matters if your problem is consistency. CoinStats gives you a powerful place to ask. BlockMind is intended to keep working even when you forget to ask.
## AI research: both products are credible [#ai-research-both-products-are-credible]
CoinStats AI Agent is not a generic chatbot attached to a tracker. Its official launch documentation says it can use on-chain data, social sentiment, technical indicators, web research, and the user's portfolio. It is available to Premium and Degen subscribers on web, iOS, and Android.
BlockMind's agent also works from portfolio context, but its strongest differentiators are persistence and judgment structure. An analysis can be reviewed by several experts rather than ending as one unqualified answer. Those verdicts can then remain part of what the agent watches.
Neither approach removes your responsibility. AI can gather evidence, expose contradictions, and make a recommendation legible. It cannot decide how much uncertainty you can afford or whether a position belongs in your life. The final call stays yours.
## Portfolio tracking and coverage [#portfolio-tracking-and-coverage]
CoinStats wins on breadth. Its current site advertises [300+ supported wallets and exchanges](https://coinstats.app/integrations/) and 120+ blockchains. Its free plan supports 10 portfolios and 20,000 transactions; Premium expands that to 100 portfolios and 100,000 transactions. It also has mature DeFi and NFT tracking.
BlockMind connects 9 exchanges through direct API sync and supports wallet tracking across 8 networks. The free classic platform allows up to 5 combined portfolios and watchlists, while Pro removes that collection limit. This is enough for many investors, but it is not a substitute for CoinStats if your holdings are spread across obscure chains, long-tail protocols, or dozens of accounts.
BlockMind's breadth runs in another direction: the agent is built to research crypto, stocks, and commodities in the same workspace. If you want one research relationship across asset classes, that matters more than the raw integration count.
## Morning Briefs and proactive monitoring [#morning-briefs-and-proactive-monitoring]
BlockMind's clearest advantage is that the output comes to you. Every morning, the agent publishes a personalized brief in the dashboard and sends it by email. The brief can cover overnight moves, portfolio-specific risks, market regime changes, and the assets or verdicts you asked it to follow.
The agent also wakes throughout the day for monitoring and alerts. This turns research from a snapshot into an ongoing responsibility: a token unlock, change in holder behavior, or broken thesis can matter weeks after the original analysis.
CoinStats offers alerts, transaction notifications, and an excellent mobile experience. Its current AI Agent and subscription documentation, however, do not list an equivalent personalized daily research brief. If you prefer opening an app and asking when you need help, that may not matter. If you want a prepared briefing waiting for you, it does.
## Mobile experience [#mobile-experience]
CoinStats is the straightforward winner for native mobile. Its iOS and Android apps carry the tracker and AI Agent, so quick checks, alerts, and research all live in a familiar mobile interface.
BlockMind is a responsive web product. It works on a phone, but it does not currently replace the polish or convenience of a mature native portfolio app. The tradeoff is that BlockMind is optimizing for a persistent agent workspace rather than a mobile-first tracker.
Choose based on the behavior you actually want. If you check a portfolio app several times a day, CoinStats fits that habit. If you want the product to prepare the context before you arrive, BlockMind is the more natural fit.
## Pricing and free tiers [#pricing-and-free-tiers]
### CoinStats [#coinstats]
CoinStats currently lists:
* **Basic:** free, with 10 portfolios, 20,000 transactions, and limited sync frequency
* **Premium:** $13.99/month billed annually after a 7-day free trial on the yearly plan, with 100 portfolios, deeper analytics, alerts, and the AI Agent
* **Degen:** a higher-capacity tier with 500 portfolios, expanded limits, and faster support
See the current [CoinStats pricing page](https://coinstats.app/pricing/) before subscribing, especially if you pay through a mobile app or outside the United States.
One published limit is inconsistent: the current pricing grid lists 100 Premium portfolios, while the FAQ on the same page says unlimited portfolios. Confirm the effective limit before purchasing if that distinction matters to you.
### BlockMind [#blockmind]
BlockMind currently offers:
* **Free classic platform:** portfolio tracking, watchlists, live charts, market indicators, and up to 5 combined portfolios and watchlists
* **Free DeepDive:** up to 5 on-demand token reports per day per email address, with no account required
* **Pro:** your dedicated agent, its workspace, Morning Briefs, research reports, expert verdicts, monitoring, and unlimited portfolios
[Current BlockMind plans, introductory terms, and the card requirement](/docs/overview/plans-and-pricing) live on the canonical pricing page so this comparison does not create a stale copy.
## Security and custody [#security-and-custody]
Both products use read-only connections for portfolio tracking. CoinStats says its integrations request read-only access, and BlockMind applies the same boundary.
For BlockMind, the trust line is simple: it uses wallet and exchange connections only to read balances and positions. The agent cannot trade, withdraw, or move funds even if you told it to. Provider-level API-key scopes vary; the [connection guide](/docs/getting-started/connect-your-portfolio) explains the Firi exception. You keep custody and make every investment decision.
Read-only access reduces risk, but it does not make any account invulnerable. Use unique credentials, enable multi-factor authentication, review API permissions, and revoke connections you no longer use.
## Which should you choose? [#which-should-you-choose]
### Choose CoinStats if you want: [#choose-coinstats-if-you-want]
* The broadest portfolio aggregation across wallets, exchanges, chains, DeFi, and NFTs
* Native iOS and Android apps
* A portfolio-aware crypto research copilot you open on demand
* A lower advertised monthly equivalent for AI research when billed annually
* Mature transaction, performance, and profit-and-loss tooling
### Choose BlockMind if you want: [#choose-blockmind-if-you-want]
* One dedicated agent with a persistent workspace
* A Morning Brief prepared around your own portfolio
* Research that moves from report to expert verdict to ongoing tracking
* Coverage across crypto, stocks, and commodities
* Proactive monitoring instead of another app you must remember to check
### Use both if: [#use-both-if]
Your holdings exceed BlockMind's integration coverage, but you still want its proactive agent. CoinStats can remain the broad tracker while BlockMind handles the portfolios and ideas that deserve deeper, ongoing attention.
## Frequently Asked Questions [#frequently-asked-questions]
### Does CoinStats have an AI agent? [#does-coinstats-have-an-ai-agent]
Yes. CoinStats launched a portfolio-aware AI research agent in April 2026. It can research markets, inspect on-chain and social data, analyze a connected portfolio, run backtests, and execute code. It is currently listed for Premium and Degen users.
### Is BlockMind cheaper than CoinStats? [#is-blockmind-cheaper-than-coinstats]
At the July 10, 2026 observation date, CoinStats advertised the lower annual entry price for its AI product. The products price different jobs: CoinStats emphasizes broad tracking and a copilot, while BlockMind Pro includes a dedicated proactive agent and workspace. Verify [BlockMind's current terms](/docs/overview/plans-and-pricing) and [CoinStats' current plans](https://coinstats.app/pricing/) before deciding.
### Which supports more wallets and exchanges? [#which-supports-more-wallets-and-exchanges]
CoinStats, by a wide margin. It advertises 300+ wallets and exchanges and 120+ blockchains. BlockMind currently supports direct sync with 9 exchanges and wallet tracking across 8 networks.
### Can either platform trade with my money? [#can-either-platform-trade-with-my-money]
BlockMind cannot trade, withdraw, or move funds. CoinStats says its portfolio-tracking connections are read-only, but it separately offers user-initiated transaction features such as swaps. Always inspect the exact permissions and action you are authorizing.
### Can I try BlockMind without starting the agent? [#can-i-try-blockmind-without-starting-the-agent]
Yes. The classic tracking platform remains free within its limits, and [DeepDive](https://blockmind.app/deepdive) provides up to 5 free token reports per day per email address without an account.
## The Bottom Line [#the-bottom-line]
CoinStats is no longer "just a tracker." It is a broad, polished crypto portfolio manager with a capable AI research copilot and strong native apps. For many crypto-only investors, it is the better value.
BlockMind earns its higher price when you want continuity: a dedicated agent that prepares a daily brief, asks several experts to challenge an idea, and keeps watching after the original research is done. Pick CoinStats for breadth and mobile convenience. Pick BlockMind for a proactive research relationship.
## Learn More [#learn-more]
* [AI Portfolio Monitoring](/blog/ai-portfolio-monitoring-crypto)
* [AI Crypto Analyst vs ChatGPT](/blog/ai-crypto-analyst-vs-chatgpt)
* [What Is an AI Crypto Agent?](/blog/what-is-an-ai-crypto-agent)
## Sources [#sources]
1. [CoinStats AI Agent launch and capability overview](https://coinstats.app/blog/ai/)
2. [CoinStats pricing and plan limits](https://coinstats.app/pricing/)
3. [CoinStats supported integrations](https://coinstats.app/integrations/)
4. [CoinStats transaction and portfolio limits](https://help.coinstats.app/en/articles/6281348-transaction-and-portfolio-limits-on-coinstats)
5. [BlockMind plans and pricing](/docs/overview/plans-and-pricing)
6. [How the BlockMind research journey works](/docs/capabilities/research-journey)
7. [BlockMind DeepDive reports](/docs/classic/deepdive-reports)
---
# Can AI Replace a Crypto Research Analyst? What an AI Agent Can (and Can't) Do
Published: 2026-03-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: informational
Primary keyword: can AI replace a crypto research analyst
Keywords: AI crypto analyst, crypto research analyst, AI investment research
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** AI can do a large part of crypto research faster than a human, especially data gathering, summarization, and pattern detection. It still does not replace human judgment on founder quality, narrative shifts, incentive design, or macro regime changes. In practice, **AI + your judgment beats either one alone**.
***
Can AI replace a crypto research analyst? **Not fully.** AI is already good enough to automate the repetitive part of research: pulling market data, scanning on-chain activity, summarizing sentiment, comparing token metrics, and flagging obvious risks. But crypto investing is not only a data problem. It is also a judgment problem. The harder questions are still human: *Does this team seem credible? Is this narrative durable or just late-cycle hype? Does the macro backdrop support risk-taking right now?*
That is why the honest answer is not “AI will replace analysts” or “AI is useless.” The honest answer is that **AI is becoming the best research assistant in crypto, but not a complete substitute for a good analyst**. Unlike a human analyst, AI does not get tired, distracted, or slow when the workload expands. Unlike a good human analyst, it still does not understand incentives, credibility, and timing in the same durable way.
## Can AI do crypto research? [#can-ai-do-crypto-research]
Yes. AI can already do a meaningful share of crypto research, especially the parts that are slow, repetitive, and data-heavy.
Crypto research has always had an ugly workload behind the scenes: reading token pages, checking liquidity, comparing market cap to fully diluted valuation, scanning holder concentration, watching whale flows, tracking exchange listings, reading community sentiment, and keeping up with changing market conditions. AI is well suited to that kind of work because it can process more inputs, more often, without getting tired.
That matters in crypto because the market is too broad for manual monitoring alone. DefiLlama reports coverage of **more than 8,000 protocols across 500+ chains**, which gives you a sense of how much surface area a serious researcher is expected to scan.[1](#sources)
Outside crypto, the case for AI-assisted analysis is already strong. In June 2025, Stanford Graduate School of Business highlighted a simulated, counterfactual AI analyst built from public information that **beat 93% of actively managed mutual fund managers over a 30-year period**.[2](#sources) That is not evidence from a deployed investor, and it does not prove AI can pick crypto winners on its own. It does show that pattern recognition across large public datasets can be powerful.
## What parts of crypto research does AI do best? [#what-parts-of-crypto-research-does-ai-do-best]
AI is best at the parts of research that depend on speed, breadth, and consistency rather than nuanced judgment.
In practice, that usually means five things.
### 1. Data gathering [#1-data-gathering]
AI can pull together market structure faster than a human can tab-switch. It can compare price, volume, volatility, support and resistance, exchange availability, holder concentration, and sentiment in one pass.
### 2. Pattern recognition [#2-pattern-recognition]
AI is good at spotting repeated setups across tokens: sudden social spikes without corresponding on-chain traction, whale concentration that looks unhealthy, momentum divergences, or a portfolio that is more exposed to one narrative than the owner realizes.
### 3. First-pass due diligence [#3-first-pass-due-diligence]
AI is very useful for turning a blank page into a shortlist. Instead of spending two hours asking “what should I even check first?”, you can have AI produce a structured first pass and then decide what deserves deeper work.
### 4. Monitoring at scale [#4-monitoring-at-scale]
Humans are bad at checking hundreds of things repeatedly. AI is good at that. The more watchlists, assets, wallets, and market indicators you follow, the bigger the advantage becomes.
This is the strongest argument for an always-on agent rather than an on-demand tool: it can watch the market while you sleep and report back every morning instead of only answering when you remember to ask.
### 5. Translation [#5-translation]
Most retail crypto holders do not need more raw data. They need plain-English interpretation. AI is good at translating messy market inputs into a clearer starting point for action.
This is where an **AI crypto analyst** is genuinely useful. Not because it magically knows the future, but because it compresses the time between “there is too much information” and “I understand what I should look at next.”
## Where does AI fall short in crypto research? [#where-does-ai-fall-short-in-crypto-research]
AI still falls short on the parts of research that depend on judgment, context, and skepticism about what the data does *not* say.
This matters more in crypto than in many other markets because crypto is full of reflexivity, hype, narrative contagion, and asymmetric information.
### 1. Team quality is not a spreadsheet field [#1-team-quality-is-not-a-spreadsheet-field]
An AI model can summarize a founder’s background, but it cannot reliably judge whether a team is credible, disciplined, promotional, evasive, or operating with the kind of integrity you would trust in a crisis. Those calls depend on reading behavior over time, not just extracting facts.
### 2. Narrative timing is partly social, not just quantitative [#2-narrative-timing-is-partly-social-not-just-quantitative]
AI can detect rising mentions, volume spikes, and sentiment shifts. It is much weaker at answering the harder question: *is this narrative early and durable, or already crowded and late?* In crypto, being right too late can lose as much money as being wrong.
### 3. Macro context changes what the same data means [#3-macro-context-changes-what-the-same-data-means]
A bullish on-chain signal during a loose-liquidity, risk-on environment does not mean the same thing during a tightening, defensive market. The IMF has argued that AI improves prediction, while humans still matter most where judgment is required.[3](#sources) That is a useful frame for crypto research too. AI can tell you what is happening in the data. You still need judgment to decide what that means *now*.
### 4. AI can sound confident when the evidence is weak [#4-ai-can-sound-confident-when-the-evidence-is-weak]
This is the biggest operational risk. A weak analyst usually looks uncertain. A weak AI output can sound polished and decisive. If you do not verify the reasoning, you can mistake fluency for insight.
### 5. Crypto scams adapt faster than templates [#5-crypto-scams-adapt-faster-than-templates]
AI can catch many obvious red flags, and that is valuable. But fraud evolves. Chainalysis estimated that crypto scams and fraud stole **at least $14 billion on-chain in 2025**, with the figure potentially exceeding **$17 billion** as more illicit addresses are identified.[4](#sources) In that environment, one templated checklist is not enough. You still need a human asking uncomfortable questions when the story looks too clean.
That is also why we recommend pairing any AI-led due diligence with basic scam awareness. If you have not already, read our guide on [5 signs of a crypto rug pull](/blog/5-signs-crypto-rug-pull).
## What does the best AI crypto analyst process look like? [#what-does-the-best-ai-crypto-analyst-process-look-like]
The best process is not “let AI decide.” It is “let AI do the heavy lifting, then apply human judgment where it matters most.”
A practical process looks like this:
1. **Use AI to gather the facts first.** Pull market data, token context, sentiment, and risk flags.
2. **Use AI to surface the important anomalies.** Look for concentration risk, weak momentum, strange holder behavior, or narrative dependence.
3. **Switch to human judgment for interpretation.** Ask whether the setup makes sense in the current macro and narrative environment.
4. **Act conservatively when conviction is low.** Size smaller, wait longer, or skip the trade.
5. **Review your own blind spots.** AI is often best at showing you what you forgot to check.
This approach also makes you less emotional. A human analyst under stress can cherry-pick data. AI can help by making the first pass more systematic. But the final call should still belong to someone who understands the cost of being wrong.
## How can BlockMind help without replacing your judgment? [#how-can-blockmind-help-without-replacing-your-judgment]
BlockMind is a **personal AI investing agent** for crypto, stocks, and commodities. On Pro, you name your agent and it gets to work in its own workspace. It knows your portfolio, scans the market daily, researches assets on demand, and sends a Morning Brief to your dashboard and email inbox. The repetitive research moves to your agent, while the judgment calls stay with you. If you want the full product overview first, read [What is BlockMind?](/docs/overview/what-is-blockmind).
Here is how that looks in practice:
* **Grounded in what you actually hold.** Connect wallets across eight supported networks and nine major exchanges. BlockMind uses those connections only to read balances and positions. The agent cannot trade, withdraw, or move funds even if you told it to. See [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
* **Research reports and expert verdicts.** [Explore](/docs/capabilities/research-journey) scans the market for ideas, Analyze turns a saved idea into an analysis report, and Verdicts asks a panel of experts for a clear call: approved, watch, wait, or reject.
* **Monitoring while you live your life.** Your agent watches your portfolio, tracked assets, and market-wide risks, and you can set [alerts in plain language](/docs/automations/monitoring-and-alerts). Every morning, its [Morning Brief](/docs/automations/morning-brief) lands on your dashboard and in your email inbox.
* **Memory that compounds.** Your agent remembers your preferences, theses, and decisions in a shared [Notebook](/docs/memory/notebook), so next month's research can build on this month's instead of starting from zero.
One boundary is deliberate: your agent does research, not financial advice. It will never tell you what to buy or sell, and it can never touch your funds. That is the right role for AI in crypto research: helping you ask better questions, notice more risks, and get to clarity faster without tempting you to outsource responsibility.
The classic platform remains free for portfolio tracking, watchlists, and live charts, with a limit of five combined portfolios and watchlists. The personal agent does not run on Free. It comes with [Pro; current plans, introductory terms, and the card requirement live on the canonical pricing page](/docs/overview/plans-and-pricing).
For a broader look at what to automate and what to judge yourself, read [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research).
## So should you trust an AI crypto analyst? [#so-should-you-trust-an-ai-crypto-analyst]
Yes, but only as a research tool, not as a substitute for judgment.
If your current process is mostly scrolling X, reacting to headlines, and checking price charts after the move already happened, AI can materially improve your research. It can make you more systematic, less reactive, and better prepared.
But if you use AI to avoid thinking, it becomes dangerous. Crypto punishes borrowed conviction. A model can help you gather evidence. It cannot carry responsibility for your decision.
The most useful mental model is simple: **AI should reduce research friction, not eliminate human accountability.**
## Frequently Asked Questions [#frequently-asked-questions]
### Can AI do crypto research? [#can-ai-do-crypto-research-1]
Yes. AI can already handle a large share of crypto research, especially data gathering, summarization, pattern recognition, and first-pass risk checks. It is much less reliable at judging founder quality, narrative durability, and macro context.
### Can AI replace a crypto research analyst? [#can-ai-replace-a-crypto-research-analyst]
Not fully. AI can replace a lot of analyst workload, but not the judgment layer that matters most in crypto. The stronger model is AI-assisted research with human decision-making.
### What is the best AI crypto analyst process? [#what-is-the-best-ai-crypto-analyst-process]
The best process is to let AI gather and structure the information first, then use human judgment to interpret it, challenge it, and decide what to do. AI should support your research, not run it end to end.
### Is BlockMind an AI crypto analyst? [#is-blockmind-an-ai-crypto-analyst]
Yes. On Pro, BlockMind gives you a personal AI investing agent that researches assets, watches your portfolio, and sends a Morning Brief to your dashboard and email inbox. The agent gathers evidence, writes analysis reports, and collects expert verdicts, but it does not make decisions for you. See [current Pro terms](/docs/overview/plans-and-pricing) on the canonical pricing page.
## The Bottom Line [#the-bottom-line]
AI is not replacing good crypto research. It is changing what “good crypto research” looks like.
The edge is no longer just who can gather the most information. The edge is who can combine fast machine-assisted analysis with calm human judgment. That is the future worth building around.
Use the [AI versus manual research framework](/blog/ai-vs-manual-crypto-research) to allocate the work, then apply the [AI-analysis verification protocol](/blog/how-to-verify-ai-crypto-analysis) before relying on a consequential answer.
## Keep reading [#keep-reading]
- [AI vs Manual Crypto Research](https://docs.blockmind.app/blog/ai-vs-manual-crypto-research)
- [What is BlockMind?](https://docs.blockmind.app/docs/overview/what-is-blockmind)
- [The research journey](https://docs.blockmind.app/docs/capabilities/research-journey)
## Sources [#sources]
1. [DefiLlama About](https://defillama.com/about), accessed July 2026.
2. [Stanford GSB: “An AI Analyst Made 30 Years of Stock Picks and Blew Human Investors Away”](https://www.gsb.stanford.edu/insights/ai-analyst-made-30-years-stock-picks-blew-human-investors-away), June 2025.
3. [IMF: “Machine Intelligence and Human Judgment”](https://www.imf.org/en/publications/fandd/issues/2025/06/machine-intelligence-and-human-judgement-ajay-agrawal), June 2025.
4. [Chainalysis: “Record $17 Billion Estimated Stolen in Crypto Scams and Fraud in 2025”](https://www.chainalysis.com/blog/crypto-scams-2026/), January 2026.
---
# Can AI Tell You When to Buy or Sell Crypto? The Honest Answer
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: problem-solving
Primary keyword: can AI tell you when to buy or sell crypto
Keywords: can AI tell you when to buy or sell crypto, AI crypto buy sell signals, AI crypto trading advice, crypto decision support, AI market timing
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** No AI can reliably tell you the perfect time to buy or sell crypto. It can help you gather current evidence, compare scenarios, identify concentration, formalize invalidation conditions, and notice when facts change. Treat any unexplained “buy now,” guaranteed return, or near-perfect win-rate claim as a risk signal—not an edge.
The appeal is obvious. Crypto trades continuously, the evidence is fragmented, and a direct answer feels more useful than a page of caveats. But exact market timing asks a model to solve three different problems simultaneously:
1. observe the relevant state without missing or stale data;
2. predict how every other participant and new event will affect price; and
3. choose an action suitable for your goals, liabilities, time horizon, and risk capacity.
Even a strong system may help with the first, model parts of the second, and know little about the third. That is why a useful AI research product should improve a decision process without issuing the decision.
The wider principle is the same one in [AI vs manual crypto research](/blog/ai-vs-manual-crypto-research): automate evidence gathering and consistency, while keeping judgment, risk, and the final decision human.
## Why “buy or sell?” is the wrong first question [#why-buy-or-sell-is-the-wrong-first-question]
A directional call hides the assumptions that matter:
* **Asset identity:** Which contract and which market?
* **Time horizon:** Five minutes, five weeks, or five years?
* **Execution:** Which venue, spread, fees, slippage, and liquidity?
* **Portfolio fit:** Is the user already concentrated in the same risk?
* **Objective:** Capital preservation, speculation, income, or long-term exposure?
* **Invalidation:** What fact would prove the thesis wrong?
Two investors can see the same evidence and rationally make different decisions. A trader with a short horizon may care about order-book liquidity and volatility. A long-horizon holder may care more about protocol adoption, supply dilution, and governance. An AI output that ignores that distinction is not personalized analysis; it is a generic opinion wearing a precise label.
## What AI can do well before a decision [#what-ai-can-do-well-before-a-decision]
### Build a current evidence pack [#build-a-current-evidence-pack]
An AI with appropriate tools can retrieve price, volume, funding, on-chain flows, governance changes, project disclosures, and news, then organize them by source and timestamp. The value is compression and consistency—not prophecy.
Require it to separate:
* verified events;
* measured market or on-chain data;
* interpretations;
* scenarios; and
* unknowns.
If the system cannot show the source, market, pair, and timestamp for a current figure, do not use that figure to make a time-sensitive decision. See [how AI accesses current crypto market data](/blog/how-ai-accesses-real-time-crypto-market-data) for the plumbing behind that rule.
### Test a thesis instead of endorsing it [#test-a-thesis-instead-of-endorsing-it]
A good prompt asks the model to attack the idea:
> My thesis is that protocol usage will increase token demand over the next 12 months. Find the strongest evidence for and against that mechanism. Distinguish product usage from token value capture. List the three observations that would invalidate the thesis.
This is more useful than “Will the token go up?” because it produces claims you can monitor. It also reduces confirmation bias: the model has an explicit job to find disconfirming evidence.
### Make portfolio consequences visible [#make-portfolio-consequences-visible]
AI can calculate concentration from connected or supplied holdings, group related exposures, and run transparent “what-if” scenarios. It can answer questions such as:
* How much of the portfolio depends on one chain, stablecoin, custodian, or narrative?
* Would an additional position duplicate an existing exposure?
* Which holdings contributed most to historical drawdown over a stated window?
* What happens arithmetically if an asset falls 20%, without claiming that it will?
Those are calculations and scenario analyses, not instructions. BlockMind’s portfolio analysis is read-only and follows the same boundary: [your agent can analyze positions but cannot trade, withdraw, or transfer](/docs/capabilities/portfolio-analysis).
### Define alerts that monitor facts [#define-alerts-that-monitor-facts]
“Tell me when to sell” is vague. Better alerts watch an observable condition:
* a governance proposal changes an admin or fee parameter;
* a known vesting wallet transfers tokens to an exchange;
* stablecoin liquidity falls below a threshold;
* price closes beyond a pre-defined zone on a stated timeframe;
* a protocol publishes a postmortem for an incident;
* a portfolio exposure exceeds a chosen percentage.
The alert says what changed and why it may matter. It still does not choose the action.
## A safer AI-assisted decision protocol [#a-safer-ai-assisted-decision-protocol]
Use this seven-step process for any consequential crypto decision.
### 1. Write the decision before researching [#1-write-the-decision-before-researching]
State the asset, intended horizon, maximum acceptable loss, and the actual question. “Should I buy?” becomes “Does the evidence support my 12-month adoption thesis strongly enough to continue manual due diligence?”
### 2. Establish the source of truth [#2-establish-the-source-of-truth]
Confirm the token contract, official documentation, relevant chain, and trading venue. Same-ticker assets and bridged representations can make a technically correct answer financially irrelevant.
### 3. Set evidence gates [#3-set-evidence-gates]
Examples:
* team and shipped product are independently verifiable;
* supply and unlock schedule reconcile across contract and disclosures;
* top-holder concentration is adjusted for exchanges and contracts;
* liquidity is sufficient for the hypothetical order size;
* the token has a plausible value-capture mechanism;
* material counterevidence has been reviewed.
Use [what to check before buying crypto](/blog/what-to-check-before-buying-crypto) for a broader data checklist.
### 4. Ask for a red-team analysis [#4-ask-for-a-red-team-analysis]
Tell the AI to assume the thesis fails and work backward. Ask which incentives, dependencies, governance powers, liquidity conditions, or hidden correlations could cause failure. Generic lists do not count; each risk should connect to evidence about this asset.
### 5. Use scenarios, not one forecast [#5-use-scenarios-not-one-forecast]
Define a base, upside, and downside case with observable drivers. Avoid precise probabilities unless you have a defensible estimation method and calibration history. “30% chance” is not more scientific because it contains a number.
### 6. Record invalidation and a review date [#6-record-invalidation-and-a-review-date]
Write what would change your mind before price moves. This prevents a falling asset from turning a testable thesis into an indefinite belief. Save the evidence, source dates, assumptions, decision, and next review.
### 7. Make the decision yourself [#7-make-the-decision-yourself]
Check whether the proposed action fits your financial situation and independent risk limits. If you need individualized financial advice, use an appropriately qualified professional—not a marketing chatbot or anonymous signal group.
## How to evaluate an AI “signal” product [#how-to-evaluate-an-ai-signal-product]
| Question | Credible behavior | Warning sign |
| --------------------------- | ------------------------------------------- | ------------------------------------------- |
| What is the output? | Evidence, uncertainty, scenarios | Unqualified buy/sell command |
| Can you inspect the method? | Inputs, time window, costs, validation | “Proprietary AI” with no details |
| Is performance testable? | Out-of-sample results, benchmark, drawdowns | Selected winning screenshots |
| Are costs included? | Fees, spread, slippage, latency | Gross return only |
| Does it discuss failure? | Regime changes and loss periods | “Works in every market” |
| Does it require custody? | Read-only research | Seed phrase, transfer, or withdrawal access |
The CFTC explicitly warns that AI cannot predict the future or sudden market changes and that claims of huge guaranteed returns or 100% win rates are used in fraud ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)). FINRA, the SEC, and NASAA likewise advise investors not to rely solely on AI-generated investment information because inputs may be outdated, incomplete, misleading, or fabricated ([FINRA](https://www.finra.org/investors/insights/artificial-intelligence-and-investment-fraud)).
## Why backtests do not settle the question [#why-backtests-do-not-settle-the-question]
A backtest can be useful evidence, but it is easy to overstate:
* **Overfitting:** the strategy was tuned to the history it is judged on.
* **Look-ahead bias:** future information accidentally entered a historical feature.
* **Survivorship bias:** failed tokens or venues were omitted.
* **Cost blindness:** spread, slippage, latency, and fees erase the apparent edge.
* **Regime dependence:** a trend strategy tested in a bull market fails in chop.
* **Selection bias:** the promoter shows the best model or period after testing many.
A result needs a pre-declared method, a realistic benchmark, out-of-sample testing, and full drawdown reporting. Even then, historical performance does not guarantee future performance.
## Limitations and counterevidence [#limitations-and-counterevidence]
AI can sometimes detect relationships in data that a person would miss. Systematic rules can also reduce impulsive decisions. Those are real advantages. They still do not prove that an AI can time a non-stationary market reliably for a new user after costs.
Current tools also face practical limits:
* Data feeds can be stale, incomplete, or venue-specific.
* Wallet labels and news explanations can be wrong.
* Social and on-chain activity may be manipulated or misclassified.
* A model can express confidence without calibration.
* The same signal can have a different meaning in a different liquidity or macro regime.
* No generic system knows your complete financial life.
## The Bottom Line [#the-bottom-line]
AI should help you ask better questions, assemble current evidence, expose contradictions, calculate portfolio consequences, and watch explicit invalidation conditions. It should not tell you what to buy or sell.
BlockMind is built around that line: your agent performs research, relates it to read-only portfolio context, and can monitor conditions, but it never makes the trade and cannot touch funds. This article is research, not financial advice.
## Sources [#sources]
* [CFTC — AI Won’t Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
* [FINRA, SEC, and NASAA — Artificial Intelligence and Investment Fraud](https://www.finra.org/investors/insights/artificial-intelligence-and-investment-fraud)
* [FINRA — Risks of Auto-Trading Services Offered by Unregistered Entities](https://www.finra.org/investors/insights/auto-trading-unregistered-entities)
* [NIST — Artificial Intelligence Risk Management Framework](https://www.nist.gov/publications/artificial-intelligence-risk-management-framework-ai-rmf-10)
---
# Does ChatGPT Know Your Crypto Portfolio? Three Ways It Can
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: commercial-investigation
Primary keyword: does ChatGPT know your crypto portfolio
Keywords: does ChatGPT know your crypto portfolio, ChatGPT crypto portfolio analysis, upload crypto portfolio to ChatGPT, AI crypto portfolio tracker, portfolio-aware AI
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** ChatGPT does not automatically know your wallets, exchange accounts, cost basis, or liabilities. It can analyze holdings you type or upload, and it can use data made available through supported connected apps or custom integrations. The result is only as current and complete as that input. For recurring portfolio monitoring, a purpose-built read-only connection reduces manual updates—but neither approach should require private keys, seed phrases, or withdrawal access.
The answer is no **by default**, not no **in principle**. Modern ChatGPT can analyze files, use apps, remember selected context, and search the web. Treating it as a permanently disconnected chatbot is out of date. Treating it as if it silently sees every wallet is equally wrong.
For the broader tool-selection landscape, read [the best AI crypto research tools](/blog/best-ai-crypto-research-tools). This article focuses specifically on portfolio context, freshness, and privacy.
## Methodology [#methodology]
Observed July 10, 2026. This comparison uses current official documentation for ChatGPT files, data analysis, apps, and privacy controls, plus BlockMind’s product documentation. It does not claim hands-on coverage testing of every wallet, exchange, app, plan, or region. The four methods are compared on input control, completeness, freshness, repeatability, permissions, and failure visibility.
The recurring workflow is covered in [AI crypto portfolio monitoring](/blog/ai-portfolio-monitoring-crypto), while [crypto portfolio concentration risk](/blog/crypto-portfolio-concentration-risk) shows one analysis a clean holdings dataset can support.
## Four levels of portfolio context [#four-levels-of-portfolio-context]
| Method | What ChatGPT can know | Freshness | Best use | Main risk |
| ------------------------- | -------------------------------------------------------- | ------------------------ | ------------------------------- | ------------------------------------- |
| Manual prompt | Holdings and facts you type | Frozen at entry time | Small one-off questions | Omission and transcription error |
| File upload | Structured balances or transaction history | Frozen at export time | Calculation, tables, charts | Stale or incomplete export |
| Connected app/integration | Data the connection exposes and you authorize | Depends on connection | Repeatable retrieval | Permissions, coverage, mapping |
| Dedicated portfolio agent | Supported read-only holdings plus saved research context | Product-specific refresh | Ongoing analysis and monitoring | Provider coverage and inferred labels |
## 1. Type a portfolio summary into the prompt [#1-type-a-portfolio-summary-into-the-prompt]
For a small portfolio, a manual table can be enough:
| Asset ID | Quantity | Location | Cost basis known? |
| -------- | -------: | ---------- | ----------------- |
| bitcoin | 0.25 | wallet A | yes |
| ethereum | 3.0 | exchange B | partial |
Then ask for arithmetic rather than advice:
> Using prices from one named source and one timestamp, calculate allocation by asset and location. Preserve unknown values, show formulas, and do not recommend trades.
This is transparent and easy to audit. It becomes fragile when you have many assets, staking positions, derivatives, multiple chains, transfers, or frequent balance changes.
### What manual input misses [#what-manual-input-misses]
* assets you forgot or considered too small;
* wallets or subaccounts not included;
* staking, lending, collateral, and borrowed balances;
* wrapped or bridged representations;
* open orders and derivatives;
* transfer history needed to avoid treating deposits as profit;
* fees, rewards, and reliable cost basis;
* current prices after the prompt was written.
The model cannot correct for a position it never received.
## 2. Upload a structured file [#2-upload-a-structured-file]
ChatGPT’s data-analysis features support spreadsheets and CSV, JSON, and other common data formats. OpenAI recommends clear column names and one record per row, and says the system can create tables, charts, calculations, and transformations ([OpenAI](https://help.openai.com/en/articles/8437071/data-analysis-with-chatgpt/)).
A useful holdings file includes:
* canonical asset ID and contract address where relevant;
* chain and wallet/account label;
* quantity and unit;
* position type: spot, staked, lent, borrowed, LP, or derivative;
* observation timestamp;
* price source and timestamp if prices are included;
* cost basis only when known and methodologically consistent;
* confidence or reconciliation status.
Ask ChatGPT to validate the file before analyzing it:
> Check duplicates, missing timestamps, negative balances, inconsistent symbols, wrapped/native double counting, and prices from mixed dates. Report issues before calculating allocation.
OpenAI notes an important constraint: the data-analysis environment cannot make external web requests. If the analysis needs external data, you must upload it or connect an available source ([OpenAI](https://help.openai.com/en/articles/8437071/data-analysis-with-chatgpt/)). A July 1 export does not become a July 10 portfolio by reasoning harder.
## 3. Use a connected app or integration [#3-use-a-connected-app-or-integration]
ChatGPT apps can let the system search or act on external services, depending on the app, permissions, account, and workspace controls. OpenAI’s current documentation says app permissions determine when ChatGPT asks before access or actions; they do not expand the access granted by the underlying connection ([OpenAI apps](https://help.openai.com/en/articles/11487775-connectors-in-chatgpt)).
A crypto portfolio connection is not automatically included merely because the app framework exists. You need a supported app or custom integration that exposes the relevant balances and transactions. Ask:
* Which wallets, exchanges, chains, and position types are covered?
* Is access truly balance/read-only?
* How often does each source refresh?
* Are liabilities and derivatives included?
* How are token identities and bridged assets normalized?
* Can the connection write, trade, transfer, or sign? If so, why?
* What happens when an account fails to sync?
For research, write access is unnecessary risk.
## 4. Use a dedicated portfolio-aware agent [#4-use-a-dedicated-portfolio-aware-agent]
A dedicated tool pre-wires portfolio ingestion, market data, and recurring research. Its advantage is not that general AI is incapable. It is that the data model, source routing, and monitoring workflow are already organized around holdings.
BlockMind’s agent can use supported connected portfolios as read-only context for allocation, exposure, performance questions, the Morning Brief, and monitoring. It cannot trade, withdraw, transfer, or move funds. The canonical connection instructions and current scope live in [Connect your portfolio](/docs/getting-started/connect-your-portfolio); the analysis boundary is documented in [Portfolio analysis](/docs/capabilities/portfolio-analysis).
That persistence changes the questions you can ask:
* “Which holdings drove today’s change?”
* “Where do I have duplicated exposure across wrappers and narratives?”
* “What changed since our last review?”
* “Which new event affects the largest position?”
* “Compare this research idea with what I already hold.”
These remain research questions, not trade instructions.
## What ChatGPT can calculate well [#what-chatgpt-can-calculate-well]
Given clean, complete data and explicit methods, ChatGPT can help with:
### Allocation [#allocation]
Calculate value weights by asset, chain, location, or category. Require one valuation timestamp and preserve missing prices.
### Concentration [#concentration]
Calculate top-position weights, Herfindahl-style concentration, or dependence on one chain, custodian, stablecoin, sector, or protocol. Category assignments are judgments; inspect them.
### Scenario arithmetic [#scenario-arithmetic]
Apply transparent hypothetical changes to positions and aggregate the result. This is not a price forecast. Avoid fake probabilities unless you have a defensible method.
### Performance reconciliation [#performance-reconciliation]
With full transactions, prices, fees, and a declared methodology, it can calculate returns. Without transfer and cash-flow history, a balance change cannot distinguish market performance from deposits and withdrawals.
### Research prioritization [#research-prioritization]
Rank which positions deserve deeper review based on exposure and material changes. Do not equate “largest” with “sell.”
## The questions portfolio AI cannot answer from balances alone [#the-questions-portfolio-ai-cannot-answer-from-balances-alone]
* What is your true risk tolerance?
* What future expenses or liabilities do you have?
* What tax treatment applies to you?
* Which wallet belongs to whom if labels are missing?
* What is your cost basis if transaction history is incomplete?
* How will correlations behave in the next crisis?
* What should you buy or sell?
Those limits are partly data problems and partly human decision problems.
## A privacy and security checklist [#a-privacy-and-security-checklist]
### Never provide secrets [#never-provide-secrets]
Do not paste a seed phrase, private key, backup code, session cookie, or unrestricted exchange credential. Public wallet addresses and balance-only connections can support research without signing authority.
### Minimize the dataset [#minimize-the-dataset]
Share only the fields needed for the question. Replace account labels with neutral aliases when identity is irrelevant. Remove names, emails, addresses, and unrelated transaction memos.
### Review data controls before uploading [#review-data-controls-before-uploading]
For individual ChatGPT services, OpenAI says content may be used to improve models depending on settings, and users can turn off “Improve the model for everyone.” Temporary Chats do not appear in history, create memories, or train models and are deleted after 30 days, subject to the documented conditions ([OpenAI data controls](https://help.openai.com/en/articles/7730893-data-usage-for-consumers)). Business products have different defaults; review the policy for your account rather than assuming.
### Inspect connection permissions [#inspect-connection-permissions]
“Read-only” should be demonstrable at the wallet, exchange, or app permission layer. Disconnect unused integrations and delete conversations or saved memory according to the provider’s documented controls.
## A reliable portfolio-analysis prompt [#a-reliable-portfolio-analysis-prompt]
> First validate the attached holdings data for duplicate assets, missing positions, inconsistent timestamps, unpriced assets, liabilities, and native/wrapped double counting. Then calculate allocation and concentration using one named price source at one timestamp. Separate observed balances, classifications, calculations, and interpretations. Show formulas, preserve unknowns, and give no buy or sell recommendation.
For recurring work, add:
> Compare with the previous snapshot, distinguish transfers from market movement where transaction data allows, and list exactly which source failed or was stale.
## Limitations and counterevidence [#limitations-and-counterevidence]
Manual ChatGPT analysis can be more controllable than an automated tracker: you choose the exact file, methodology, and moment, and can inspect generated analysis code. A dedicated connection can introduce mapping errors or coverage gaps invisibly.
Conversely, manual files often go stale and omit small or complex positions. Neither method is intrinsically superior. Reliability depends on completeness, timestamps, permissions, reconciliation, and whether the output exposes assumptions.
Memory is also not a ledger. Even if a system remembers that you hold an asset, do not assume it remembers the exact live quantity or reconciles every change.
## The Bottom Line [#the-bottom-line]
ChatGPT knows only the portfolio context you explicitly provide or authorize through a supported connection. For one-off analysis, a clean file and explicit methodology can work well. For ongoing monitoring, a purpose-built read-only agent reduces repeated data preparation. In both cases, verify completeness, prices, timestamps, and privacy controls.
BlockMind’s portfolio features are research, not financial advice. Your agent never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [OpenAI — Data Analysis with ChatGPT](https://help.openai.com/en/articles/8437071/data-analysis-with-chatgpt/)
* [OpenAI — Apps in ChatGPT](https://help.openai.com/en/articles/11487775-connectors-in-chatgpt)
* [OpenAI — Data Controls FAQ](https://help.openai.com/en/articles/7730893-data-usage-for-consumers)
* [OpenAI — How Your Data Is Used to Improve Model Performance](https://help.openai.com/en/articles/5722486-chatgpt-privacy-policies)
* [BlockMind — Connect Your Portfolio](/docs/getting-started/connect-your-portfolio)
---
# Can ChatGPT Give Accurate Crypto Prices? What to Verify
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Tools and Comparisons
Search intent: problem-solving
Primary keyword: can ChatGPT give accurate crypto prices
Keywords: can ChatGPT give accurate crypto prices, ChatGPT crypto prices, ChatGPT Bitcoin price accuracy, AI real-time crypto price, verify crypto price
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** ChatGPT can return a current and accurate crypto price when it searches the web or uses an appropriate connected source. It can also return a stale, approximate, differently denominated, or wrong-asset number. A usable quote must identify the asset or contract, venue or aggregator, trading pair, price type, source timestamp, and retrieval time. For execution, verify directly on the venue you will use.
The outdated answer is “ChatGPT has no internet access.” OpenAI’s current documentation says ChatGPT can search the web and cite sources when it needs current information ([OpenAI](https://help.openai.com/en/articles/12677804-what-is-chatgpt-faq)). Search solves access; it does not make “the price” a single unambiguous fact.
If you are choosing an entire research stack rather than verifying one quote, start with [the best AI crypto research tools](/blog/best-ai-crypto-research-tools). This guide owns the narrower question of price identity and freshness.
## Evidence scope (July 10, 2026) [#evidence-scope-july-10-2026]
This guide relies on current official OpenAI capability documentation and official market-data API documentation. We did not run or claim a controlled ChatGPT price-accuracy test. Results can vary by plan, enabled tools, prompt, source availability, asset, and moment; the verification protocol below is designed to evaluate an individual answer without assuming a universal error rate.
For the source-to-answer plumbing behind freshness, see [how AI accesses real-time crypto market data](/blog/how-ai-accesses-real-time-crypto-market-data). For deterministic quote and threshold use cases, compare an [AI crypto tool with a price bot](/blog/ai-crypto-tool-vs-price-bot).
## Why two correct crypto prices can disagree [#why-two-correct-crypto-prices-can-disagree]
Suppose two sources report different BTC prices at the same moment. Neither must be broken. They may differ because of:
* **Venue:** Coinbase, Kraken, a decentralized exchange, and another market have separate order books.
* **Pair:** BTC/USD, BTC/USDT, BTC/EUR, and a synthetic or derivative contract are different instruments.
* **Price type:** last trade, bid, ask, midpoint, mark price, index price, or daily close.
* **Timing:** one source streams trades; another refreshes every several seconds.
* **Aggregation:** an index may combine venues and exclude outliers.
* **Order size:** a displayed reference price is not the average price for a large market order.
* **Asset identity:** tickers can collide, and bridged or wrapped versions can trade separately.
Accuracy therefore means “correct for a defined instrument, source, and moment,” not “matches every screen.”
## How ChatGPT gets a current number [#how-chatgpt-gets-a-current-number]
### Web search [#web-search]
ChatGPT can search for a recent result and cite the page. This is convenient for general research. The result may still come from a cached snippet, a page with its own refresh cadence, or an aggregator that does not describe executability.
### A connected app or custom integration [#a-connected-app-or-custom-integration]
Depending on account and workspace capabilities, ChatGPT apps can connect to external services and data. OpenAI’s documentation says permissions and the access granted to each app determine what data and actions are available ([OpenAI apps](https://help.openai.com/en/articles/11487775-connectors-in-chatgpt)). A market-data integration can return structured values with clearer provenance than open-web search, but its quality still depends on the integration.
### User-supplied data [#user-supplied-data]
ChatGPT can analyze uploaded CSV, spreadsheet, JSON, and text data. OpenAI notes that the data-analysis environment cannot make external web requests, so external data must be uploaded or connected first ([OpenAI data analysis](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt/)). A portfolio export or price file is a snapshot, not a self-refreshing feed.
### Model memory [#model-memory]
Without current retrieval, a model may answer from learned patterns or prior conversation context. That is unsuitable for a time-sensitive quote even if the number happens to look plausible.
## The seven fields in a trustworthy crypto quote [#the-seven-fields-in-a-trustworthy-crypto-quote]
Ask for this format:
> Report the exact asset and contract if relevant; venue or aggregator; trading pair; last, bid/ask, midpoint, mark, index, or close; value; source timestamp; retrieval timestamp; and a direct source link. If any field is unavailable, say so.
### 1. Canonical asset identity [#1-canonical-asset-identity]
For BTC and ETH, a name may be enough for casual research. For smaller tokens, require chain and contract address. Tickers are not unique.
### 2. Source [#2-source]
Name the exchange or aggregator. “According to the web” cannot be audited.
### 3. Pair and denomination [#3-pair-and-denomination]
A USD quote and a stablecoin quote are not guaranteed to be identical. If conversion is involved, identify the conversion source and time.
### 4. Price type [#4-price-type]
Last trade is not the same as the current bid or ask. A derivatives mark is designed for a different purpose than a spot last trade. A daily close depends on venue and cutoff.
### 5. Source timestamp [#5-source-timestamp]
This tells you when the provider says the value was current. CoinGecko’s simple-price response can include `last_updated_at`, and its documentation states a 20-second cache/update frequency for listed Pro plans ([CoinGecko](https://docs.coingecko.com/reference/simple-price)).
### 6. Retrieval timestamp [#6-retrieval-timestamp]
This tells you when ChatGPT or the integration fetched the source. It does not replace the source timestamp.
### 7. Executability [#7-executability]
If the purpose is a trade, a reference quote is insufficient. Check the venue’s bid/ask, depth, fees, and estimated slippage for the intended order size. A correct midpoint can still be impossible to execute.
## When ChatGPT is accurate enough [#when-chatgpt-is-accurate-enough]
ChatGPT search is often suitable for:
* explaining roughly where a major asset trades while citing the source;
* comparing current market-cap estimates with clear timestamps;
* finding official API documentation;
* summarizing a historical period after defining the source and close convention;
* translating a structured market-data response into plain language.
The risk is proportional to the decision. A seconds-old aggregated quote may be perfectly adequate for an educational explanation and inadequate for liquidation-risk monitoring.
## When to use a dedicated source instead [#when-to-use-a-dedicated-source-instead]
Use the exchange or a structured market-data provider when you need:
* order entry or execution verification;
* bid, ask, depth, spread, or expected price impact;
* high-frequency alerts;
* derivatives mark and liquidation data;
* a reproducible historical series;
* one defined methodology across many assets;
* service-level expectations for data delivery.
Coinbase documents a public WebSocket feed for real-time orders and trades and describes sequence numbers for detecting dropped or out-of-order messages ([Coinbase](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)). That is the kind of source a production price system consumes. A chat search result is a different tool.
## A five-minute verification test [#a-five-minute-verification-test]
You do not need to “trust ChatGPT” or reject it categorically. Test the answer.
1. Ask for the seven fields above.
2. Open the cited source rather than relying on the snippet.
3. Check that the asset and contract match.
4. Compare the timestamp and pair with a second independent primary market source.
5. If the value matters for execution, inspect the actual venue’s order book.
Record discrepancies by category. A different venue is not the same failure as an invented source. This approach evaluates the quote rather than the confidence of the prose.
## Better prompts for common price questions [#better-prompts-for-common-price-questions]
### Current price [#current-price]
> What is the latest BTC/USD price? Use a named source, identify price type, include source and retrieval timestamps, and link directly to the source. Do not use model memory.
### Smaller token [#smaller-token]
> Resolve TOKEN by chain and contract address first. Then report its USD price from two sources, liquidity or volume context, and timestamps. Stop if identity is ambiguous.
### Historical price [#historical-price]
> What was the daily closing price on YYYY-MM-DD? State venue, pair, timezone, candle interval, and whether the value is adjusted or aggregated.
### Portfolio valuation [#portfolio-valuation]
> Value each supplied balance using one named pricing methodology at the same timestamp. Preserve missing prices instead of substituting zero, and flag stale or illiquid assets.
### Price-move explanation [#price-move-explanation]
> Confirm the move from a defined source first. Then separate verified contemporaneous events from plausible explanations. Do not imply causation merely because a headline appeared nearby.
## Common failure modes [#common-failure-modes]
* **Ticker collision:** the answer prices the wrong token.
* **Stale snippet:** the visible number was cached before retrieval.
* **Pair mismatch:** USD and USDT are treated as identical.
* **Market mismatch:** spot, perpetual, and index prices are blended.
* **Timestamp omission:** a historical number is presented as current.
* **False precision:** many decimals disguise an approximate input.
* **Circular sourcing:** several pages repeat one upstream quote.
* **Cause invention:** the answer attaches price movement to the nearest news item.
* **No-liquidity quote:** a tiny last trade is treated as a scalable valuation.
OpenAI’s own accuracy guidance states that ChatGPT can produce incorrect facts and fabricated citations and advises checking important information, even when search tools can improve recency ([OpenAI](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)).
## Limitations and counterevidence [#limitations-and-counterevidence]
Dedicated crypto tools are not automatically correct. Exchanges can experience outages, API clients can miss messages, aggregators can mis-map tokens, and data providers can revise methodology. A direct API value without identity and timestamp can be less useful than a well-cited search answer.
ChatGPT can also be the better interface when the job is explanation rather than low-latency execution. The point is not that general AI is stale. The point is that every current price is a sourced observation with a scope.
## The Bottom Line [#the-bottom-line]
Yes, ChatGPT can give an accurate crypto price. Verify what the number means. Require asset identity, source, venue or aggregation, pair, price type, source time, retrieval time, and liquidity context. For a trade, the final source of truth is the venue and instrument you will actually use.
BlockMind can combine current market research with read-only portfolio context, but it is a research product—not a trading venue or financial adviser. Your agent never tells you what to buy or sell and cannot touch funds.
This article is research, not financial advice.
## Sources [#sources]
* [OpenAI — What Is ChatGPT?](https://help.openai.com/en/articles/12677804-what-is-chatgpt-faq)
* [OpenAI — Does ChatGPT Tell the Truth?](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)
* [OpenAI — Data Analysis with ChatGPT](https://help.openai.com/en/articles/8437071/data-analysis-with-chatgpt/)
* [CoinGecko API — Simple Price Endpoint](https://docs.coingecko.com/reference/simple-price)
* [Coinbase Developer Documentation — Exchange WebSocket Overview](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)
---
# Crypto Alerts Beyond Price: What to Monitor and How to Control Noise
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: problem-solving
Primary keyword: crypto alerts beyond price
Keywords: crypto alerts beyond price, crypto portfolio alerts, on-chain alerts, crypto risk alerts, wallet monitoring alerts
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** The most useful crypto alerts often happen before or outside price: large wallet flows, liquidity removal, admin or proxy changes, token unlocks, stablecoin stress, protocol incidents, custody problems, and thesis-breaking news. Each alert needs a trusted source, a precise trigger, an expected latency, and a predefined research response. Alerts inform; they should not trade for you.
***
Price is a summary. It rarely explains what changed.
A protocol can upgrade before its token moves. A large holder can transfer tokens before any sale. Liquidity can be removed while a quoted price still looks stable. An exchange can pause withdrawals while portfolio value remains unchanged. A thesis can deteriorate gradually through lost users or missed milestones without crossing a price line.
The portfolio-intelligence pillar, [AI Portfolio Monitoring for Crypto](/blog/ai-portfolio-monitoring-crypto), explains the broader monitoring job. This guide turns that job into an event taxonomy and a noise-controlled alert design.
## Why price-only alerts are incomplete [#why-price-only-alerts-are-incomplete]
Price alerts are useful because they are simple, comparable, and widely available. CoinGecko's [official guide](https://www.coingecko.com/learn/coingecko-crypto-price-alerts-portfolio) documents one-time and recurring target-price notifications. But one price can hide several causes:
* broad market movement;
* thin-liquidity trade;
* exploit or governance event;
* insider unlock or distribution;
* exchange listing or delisting;
* stablecoin or bridge stress;
* false rumor or coordinated promotion.
The price threshold says “look now.” It does not say what to look at.
## Seven alert layers [#seven-alert-layers]
### 1. Portfolio movement alerts [#1-portfolio-movement-alerts]
Track changes that are material *to your holdings*, not only a universal token threshold:
* portfolio drawdown over a defined period;
* one position's contribution to the move;
* position weight drifting from a reference range;
* unexplained balance change;
* new asset, debt, or spam token appearing;
* exchange sync or wallet-scan failure.
Start with accurate data. The [crypto portfolio health check](/blog/crypto-portfolio-health-check) explains reconciliation and look-through exposure. A bad balance creates bad alerts.
### 2. Wallet and flow alerts [#2-wallet-and-flow-alerts]
Block explorers and on-chain platforms can notify on address activity. [Etherscan's Watch List](https://info.etherscan.com/watch-list/) sends email when selected addresses have incoming or outgoing transactions. [Nansen Smart Alerts](https://academy.nansen.ai/en/help/articles/6239622-ai-smart-alerts-101) supports addresses, entities, token flows, exchange flows, and contract interactions.
Useful targets include:
* deployer and treasury wallets;
* vesting and unlock contracts;
* protocol multisigs;
* large holders relevant to a thesis;
* bridge, pool, and locker contracts;
* your own public addresses;
* exchange-associated addresses, with attribution caveats.
A transfer is not a sale. It can be custody rotation, collateral movement, staking, bridging, market-making inventory, or internal accounting. The alert opens an investigation.
### 3. Liquidity alerts [#3-liquidity-alerts]
Monitor the market's ability to absorb trades:
* large LP-token or position-NFT transfer;
* liquidity lock nearing expiry;
* large removal from the active pool;
* active v3/v4 range moving away from price;
* depth falling below a research threshold;
* spread or slippage widening;
* trading migrating to an unfamiliar pool.
Use the exact pool and position from the [liquidity locks guide](/blog/liquidity-locks-crypto-explained). A token-level alert can miss the pool that matters.
### 4. Contract and governance alerts [#4-contract-and-governance-alerts]
Watch powers that can change token or protocol behavior:
* proxy implementation upgrade;
* admin, owner, minter, pauser, or upgrader role change;
* tax, blacklist, whitelist, maximum transaction, or trading-status change;
* multisig signer or threshold change;
* timelock proposal queued or executed;
* emergency pause;
* material governance proposal or vote result.
These events may be legitimate maintenance. The alert should link to the transaction, proposal, diff, audit, and official explanation.
### 5. Tokenomics and holder alerts [#5-tokenomics-and-holder-alerts]
* scheduled unlock or vesting release;
* large receiving-wallet movement after unlock;
* supply mint or burn;
* top-holder concentration change;
* treasury runway or emissions change;
* governance proposal changing incentives;
* bridge supply divergence.
Do not alert on every transfer. Group known vesting addresses and set materiality relative to circulating supply and pool depth.
### 6. Protocol, stablecoin, and custody alerts [#6-protocol-stablecoin-and-custody-alerts]
* official exploit or incident notice;
* abnormal outflow from critical contracts;
* oracle, bridge, sequencer, validator, or chain halt;
* stablecoin price deviation plus reserve/redemption news;
* exchange withdrawal pause or status incident;
* custody-provider security disclosure;
* audit or vulnerability publication.
Require at least one primary operational source where possible: protocol status page, issuer, exchange, governance forum, repository, or on-chain transaction. Social posts can provide speed; they need verification.
### 7. Thesis alerts [#7-thesis-alerts]
The hardest alerts are qualitative:
* roadmap milestone missed or re-scoped;
* usage, fees, users, or developer activity diverge from the thesis;
* competitor changes the market structure;
* team or governance loses key contributors;
* regulatory treatment changes;
* a claimed partnership is contradicted;
* a risk previously labeled “temporary” persists.
Turn each thesis into a falsifiable sentence and attach observable evidence. “Alert me if the project is bad” is not a useful rule. “Recheck if the production launch moves beyond Q3 or the official roadmap changes” is auditable.
## The STACK alert design rubric [#the-stack-alert-design-rubric]
Every alert should define five fields:
### S — Source [#s--source]
Where does the event come from? Prefer contract events, official status pages, issuer disclosures, exchange APIs, and primary documents. Record backup sources.
### T — Trigger [#t--trigger]
What exact condition must be true? Include asset or address, chain, field, direction, threshold, window, and whether the alert repeats.
### A — Attribution [#a--attribution]
What does the signal *not* prove? State plausible alternatives. For example, “exchange inflow” is an address-label inference, not confirmed sale intent.
### C — Cadence and latency [#c--cadence-and-latency]
How often is the source checked, and how late can delivery be? Do not call a periodic scan real-time. Match the tool to the urgency.
### K — Known response [#k--known-response]
What research step follows? Examples: open the transaction, verify the official notice, reconcile the wallet, compare pool depth, or update the thesis. Avoid an automatic buy or sell response unless you have deliberately chosen and controlled an execution system outside this research framework.
## Set a notification noise budget [#set-a-notification-noise-budget]
Estimate daily notification load before activating a rule:
`expected alerts per day = monitored entities × event rate per entity × fraction passing filters`.
**Hypothetical example:** You monitor 20 wallets. They average 3 transfers per day, and a $50,000 materiality filter plus exclusions lets 5% through. Expected load is `20 × 3 × 0.05 = 3 alerts per day`.
If you also add 15 protocol and market rules expected to fire 0.2 times per day each, that contributes `15 × 0.2 = 3` more. The total is about six daily alerts.
This is a planning estimate, not a guarantee. Event rates cluster during stress. If six is too many to investigate, raise materiality, group related events, use a summary, or split urgent and review-later channels.
## A severity and response table [#a-severity-and-response-table]
| Severity | Example | Delivery target | Research response |
| -------- | ------------------------------------------------------------- | ------------------------- | ------------------------------------------------ |
| Critical | Wallet compromise indicator, exploit, unexpected admin change | Fastest verified channel | Verify source, secure access, inspect exposure |
| High | Large liquidity removal, stablecoin stress, withdrawal pause | Prompt notification | Reconcile affected positions and primary updates |
| Medium | Unlock movement, concentration change, thesis milestone slip | Same-day queue | Update evidence ledger |
| Low | Routine governance, minor balance drift, weekly metric change | Brief or scheduled review | Batch with related context |
Severity should reflect impact on *your exposure*, confidence in the event, and time sensitivity—not how dramatic the headline sounds.
## BlockMind alerts: the honest boundary [#blockmind-alerts-the-honest-boundary]
BlockMind's agent runs monitoring passes every few hours during the documented daily window. Default passes cover large portfolio moves, tracked-asset moves and major news, plus wider market risks such as stablecoin depegs, exploits, governance emergencies, and bridge halts.
Users can set durable plain-language rules for absolute price, 24-hour percentage change, and asset ratios, and can ask for recurring research checks such as reviewing funding rates. Alerts can arrive in-app, by email, or through Telegram when connected. The exact behavior is documented in [Monitoring and alerts](/docs/automations/monitoring-and-alerts).
This is periodic research monitoring, not second-by-second execution. It may take a few hours to evaluate and deliver a condition. For wallet-event or contract-event immediacy, a chain explorer or specialist on-chain alert tool may be the appropriate source. BlockMind cannot trade or move funds.
## A copyable alert specification [#a-copyable-alert-specification]
```text
Name:
Research question:
Entity/address/asset:
Chain and exact identifier:
Primary source:
Trigger and window:
Materiality filter:
Known exclusions:
Maximum acceptable latency:
Repeat or one-time:
Delivery channel:
What the signal does NOT prove:
First verification step:
Evidence ledger to update:
Expiry/review date:
```
Review rules after migrations, address changes, strategy changes, or repeated false positives.
## Alert checklist [#alert-checklist]
* [ ] Exact asset, address, pool, or contract recorded
* [ ] Primary and backup sources chosen
* [ ] Trigger includes direction, amount, and time window
* [ ] Spam, internal transfers, and system addresses filtered
* [ ] Latency matches urgency
* [ ] Expected alert load estimated
* [ ] Severity and channel assigned
* [ ] Alternative explanations written down
* [ ] Verification response defined
* [ ] Rule has an owner and review date
* [ ] No seed phrase, private key, or excessive API permission provided
* [ ] Alert informs rather than silently executes
## Limitations and counterevidence [#limitations-and-counterevidence]
Alerts fail in both directions. False positives create panic and fatigue; false negatives create misplaced confidence. Labels can be wrong, sources can lag, APIs can fail, and critical events can happen between periodic checks. Attackers may split transactions below thresholds or use new addresses.
More alerts are not more safety. A small number of thesis-linked rules that someone will actually investigate is usually more useful than an unowned stream of raw events.
## The Bottom Line [#the-bottom-line]
Price alerts tell you that the market moved. A complete monitoring system asks whether portfolio data, wallets, liquidity, contracts, token supply, protocols, custody, or the thesis changed first.
Design every rule with STACK and keep a noise budget. The output of an alert is a research task—not a trading instruction.
This article is for research and education, not financial or security advice.
## Sources [#sources]
1. [Etherscan: Address Watch List notifications](https://info.etherscan.com/watch-list/)
2. [Nansen: AI Smart Alerts documentation](https://academy.nansen.ai/en/help/articles/6239622-ai-smart-alerts-101)
3. [CoinGecko: Portfolio and price-alert guide](https://www.coingecko.com/learn/coingecko-crypto-price-alerts-portfolio)
4. [OpenZeppelin: Access-control roles and timelocks](https://docs.openzeppelin.com/contracts/5.x/access-control)
5. [Uniswap: Liquidity-position ownership](https://support.uniswap.org/hc/en-us/articles/20980786685069-Why-is-liquidity-position-ownership-represented-by-tokens-or-NFTs)
6. [BlockMind: Monitoring and alerts](/docs/automations/monitoring-and-alerts)
7. [BlockMind: Portfolio analysis](/docs/capabilities/portfolio-analysis)
---
# Crypto Community Analysis: Measure Substance, Not Follower Counts
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: analyze a crypto community
Keywords: crypto community analysis, evaluate crypto community, crypto community red flags, Discord Telegram crypto analysis, crypto social media due diligence
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A large crypto audience is not necessarily a healthy community. Measure whether members can get accurate answers, challenge claims without punishment, participate in governance and development, retain activity after incentives end, and distinguish official communication from spam. Sample conversations and outcomes across multiple weeks; do not turn followers, members, or message counts into evidence of product demand.
***
## What is crypto community analysis? [#what-is-crypto-community-analysis]
Crypto community analysis evaluates the quality, authenticity, resilience, and productive output of the people and processes around a project.
It asks six different questions:
1. **Support quality:** Can users solve real problems?
2. **Information quality:** Are claims sourced, corrected, and current?
3. **Critical tolerance:** Can members ask difficult questions?
4. **Participation:** Do members contribute code, governance, education, or operations?
5. **Retention:** Does meaningful activity persist without campaigns and price spikes?
6. **Integrity:** How much activity appears automated, coordinated, impersonated, or incentive-driven?
Follower count answers none of those directly.
**Scope and assumptions:** This method covers public forums, GitHub, governance portals, Discord, Telegram, X, Reddit, and project calls. Private channels and deleted messages create unavoidable gaps. Do not collect personal data beyond what is necessary for public research, and do not attempt to deanonymize members.
*This is research, not financial advice. BlockMind will not tell you what to buy or sell, and its agent cannot touch funds.*
Community quality is one input to the broader [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto); it cannot substitute for code, tokenomics, liquidity, or security evidence.
## Why community metrics are easy to manipulate [#why-community-metrics-are-easy-to-manipulate]
Social platforms face automated and coordinated behavior at scale. Discord defines spam to include mass account creation or use to distribute unwanted or malicious content and describes generated, compromised, and human-operated spam accounts ([Discord Safety](https://discord.com/safety/how-discord-is-fighting-spam)).
Crypto adds direct financial incentives:
* Airdrop eligibility
* Referral rewards
* Ambassador points
* Giveaway entries
* Token-gated roles
* Paid promotion
* Price-sensitive treasury or token holdings
The CFTC warns that pump-and-dump organizers use social media and messaging apps to hype thinly traded tokens, sometimes with false news and coordinated urgency ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)). Therefore activity volume must be separated from independent, useful participation.
## The community substance matrix [#the-community-substance-matrix]
| Dimension | Evidence to sample | Stronger pattern | Warning pattern |
| ------------------ | --------------------------------- | -------------------------------------- | ------------------------------------ |
| Support | User questions and resolution | Specific, timely, documented answers | Questions ignored or pushed to DMs |
| Information | Announcements and corrections | Primary links, changelogs, corrections | Unsourced claims, deleted errors |
| Critical tolerance | Hard questions | Direct answers and civil disagreement | Bans, ridicule, price-only replies |
| Governance | Proposals, delegates, execution | Deliberation before voting | Rubber-stamp votes or opaque control |
| Contribution | Code, docs, translations, tooling | Output reaches product/community | Role farming without output |
| Retention | Cohort activity over time | Meaningful repeat participation | Spikes only near rewards or price |
| Integrity | Account and message patterns | Diverse language and behavior | Copy-paste waves, impersonation |
| Safety | Official links and moderation | Clear reporting and scam warnings | Admin DMs, unsafe links, no policy |
Preserve examples and dates behind each label.
## Step 1: map channel purpose and ownership [#step-1-map-channel-purpose-and-ownership]
Build the canonical map from the official project site:
| Channel | Official link verified? | Purpose | Who controls it? | Archive/search available? |
| ---------------- | ----------------------- | ---------------------- | ---------------- | ------------------------- |
| Governance forum | | Proposals/deliberation | | |
| GitHub | | Code/issues | | |
| Discord | | Support/community | | |
| Telegram | | Announcements/chat | | |
| X | | Public announcements | | |
| Calls | | Updates/governance | | |
Do not treat unofficial price groups as the project community. Confirm admin identities through official links, not direct messages.
Discord advises users that official announcements should come through official channels and warns against unfamiliar links and files ([Discord Safety](https://discord.com/safety/360044104071-tips-against-spam-and-hacking)). Apply the same standard to research: navigate from the project’s official domain and preserve the route.
## Step 2: use a time-stratified sample [#step-2-use-a-time-stratified-sample]
Do not sample only the last 100 messages. That can capture one event or coordinated campaign.
Choose at least four windows:
* Ordinary week
* Product release or governance week
* Material incident or market-stress week
* Incentive/airdrop or price-spike week
Within each window, sample the same types of evidence:
* First 25 substantive user questions
* First 10 critical or skeptical questions
* All official corrections
* Five governance discussions
* Five contributor pathways from proposal to output
State the channel, dates, inclusion rules, and sample size. This is not a statistically representative survey unless designed as one; call it a structured qualitative sample.
## Step 3: measure support resolution [#step-3-measure-support-resolution]
For sampled user questions, record:
```text
Resolution rate = questions with a verifiable resolution / substantive questions sampled
```
Also record:
* Time to first useful response
* Whether a maintainer, moderator, or peer answered
* Whether the answer links current official documentation
* Whether the issue was escalated to GitHub or a status page
* Whether the user confirmed resolution
* Whether recurring questions produced documentation improvements
Avoid universal “good” thresholds. Complexity and volunteer coverage differ. Compare the project with its own published support model and with similar projects.
## Step 4: audit information quality [#step-4-audit-information-quality]
Select material claims about:
* Releases
* Partnerships
* Token supply and unlocks
* Security and audits
* User or transaction counts
* Governance outcomes
* Listings
For each claim, verify the primary source. Use the same status system as [How to Read a Crypto Whitepaper](/blog/how-to-read-crypto-whitepaper): verified, partly verified, unverified, contradicted, or stale.
Measure correction behavior:
* Was the original error retained with a correction?
* Was the corrected source linked?
* Did all official channels update?
* Did moderators stop repeated misinformation?
A community that corrects itself visibly can be more trustworthy than one that appears error-free because mistakes disappear.
## Step 5: test critical tolerance [#step-5-test-critical-tolerance]
Look for questions that challenge:
* Treasury spending
* Admin keys
* Token unlocks
* Failed roadmap items
* Security incidents
* Validator or delegate concentration
* Partnerships and metrics
Code responses:
| Response | Description |
| --------- | -------------------------------------------------- |
| Direct | Answers the claim with evidence |
| Partial | Addresses some elements and acknowledges gaps |
| Deferred | Names an owner and follow-up date |
| Deflected | Changes topic, attacks motive, or repeats slogans |
| Removed | Deleted or sanctioned without visible policy basis |
Moderation is not automatically censorship. Spam, harassment, scams, and doxxing should be moderated. The question is whether good-faith evidence requests are handled under clear, consistently enforced rules.
GitHub’s code-of-conduct guidance emphasizes defining engagement standards and procedures for addressing abuse ([GitHub Docs](https://docs.github.com/en/communities/setting-up-your-project-for-healthy-contributions/adding-a-code-of-conduct-to-your-project)). A documented policy helps separate legitimate moderation from arbitrary suppression.
## Step 6: connect community participation to output [#step-6-connect-community-participation-to-output]
Follow participation chains:
```text
Question or idea
→ issue/proposal
→ discussion and owner
→ implementation or decision
→ release/execution
→ documented outcome
```
Examples of productive output:
* Code merged into canonical repositories
* Documentation corrected
* Governance proposals executed
* Independent dashboards or research reproduced
* Local events with public materials
* Translations maintained
* Vulnerabilities responsibly disclosed
* Support questions converted into knowledge-base entries
Use [Crypto GitHub Developer Activity](/blog/crypto-github-developer-activity) to verify code contributions and releases.
## Step 7: analyze governance participation correctly [#step-7-analyze-governance-participation-correctly]
Raw voter count or token turnout does not reveal governance quality.
Inspect:
* Proposal creation requirements
* Delegation concentration
* Quorum and approval thresholds
* Discussion period before voting
* Late changes to proposals
* Voter overlap and conflicts
* Execution after approval
* Admin or guardian vetoes
* Treasury-recipient disclosure
Calculate concentration when data allows:
```text
Top-10 voting share = voting power of top 10 participating entities / total participating power
```
Address count is not entity count. Delegates may represent many holders, and one entity can use several addresses.
## Step 8: identify manipulation and incentive artifacts [#step-8-identify-manipulation-and-incentive-artifacts]
Warning patterns include:
* Large batches of new accounts posting identical phrases
* Engagement concentrated around giveaways
* Replies unrelated to the original post
* Repeated price targets and urgency
* Accounts active only for campaign tasks
* Sudden member spikes without product or event explanation
* Impersonated admins asking for funds or seed phrases
* High message count with very few unique substantive questions
* Referral links dominating support channels
* Critical messages flooded by copy-paste positivity
Do not accuse individuals of being bots from one pattern. Label the behavior: “coordinated-looking copy-paste activity” or “incentive-linked accounts,” and explain the evidence.
## Worked hypothetical: “Lattice Protocol” [#worked-hypothetical-lattice-protocol]
**Lattice is fictional. This is not a real community assessment.**
Lattice advertises 180,000 Discord members and 400,000 X followers.
A four-window sample finds:
* Ordinary weeks average 35 substantive Discord questions; 24 receive verifiable resolutions.
* Release week has 120 substantive questions and 63 resolutions; unresolved issues are moved to GitHub, but only half receive follow-up links.
* Giveaway week message volume rises 9×, while unique technical questions rise only 15%.
* Of 50 sampled giveaway messages, 32 use one of four near-identical phrases.
* Five difficult treasury questions receive direct links to spending reports; two questions about an admin key are repeatedly deferred with no owner.
* Three community proposals are approved; two execute, while one has no status update after eight weeks.
* Four external code contributors merge changes, but one maintainer reviews 91% of external pull requests.
Responsible conclusion:
> Lattice has a functional support and contributor core, plus transparent treasury responses. Headline audience and message counts are heavily affected by incentives. Admin-key disclosure and proposal follow-through are weak, and review capacity is concentrated.
The output preserves both supportive and adverse evidence.
## A reproducible community review template [#a-reproducible-community-review-template]
```text
SCOPE
- Official channels and verification path:
- Sampling windows:
- Sample sizes and exclusions:
SUPPORT
- Substantive questions sampled:
- Resolution rate:
- Escalation and documentation pattern:
INFORMATION
- Material claims checked:
- Verified / partial / unknown / contradicted / stale:
- Correction examples:
CRITICAL TOLERANCE
- Hard questions sampled:
- Direct / partial / deferred / deflected / removed:
- Moderation policy:
PARTICIPATION
- Proposals to execution:
- Contributions to release:
- Documentation/tooling outcomes:
RETENTION AND INTEGRITY
- Ordinary vs event activity:
- Incentive-linked activity:
- Coordinated-looking patterns:
GOVERNANCE
- Turnout and concentration:
- Deliberation and execution:
- Guardian/admin controls:
CONCLUSION
- Strongest evidence:
- Strongest counterevidence:
- Unknowns and next review trigger:
```
## Common crypto community analysis mistakes [#common-crypto-community-analysis-mistakes]
### Equating follower count with users [#equating-follower-count-with-users]
Followers can include inactive, purchased, duplicated, incentivized, or curious accounts. Product usage needs separate evidence.
### Equating message volume with health [#equating-message-volume-with-health]
Spam, price chatter, and campaigns can dominate volume. Sample substance and outcomes.
### Treating criticism as community weakness [#treating-criticism-as-community-weakness]
Evidence-based disagreement and visible correction can be signs of resilience.
### Treating moderation as proof of manipulation [#treating-moderation-as-proof-of-manipulation]
Healthy communities enforce safety rules. Inspect whether policies are clear and consistently applied.
### Sampling only bullish or crisis periods [#sampling-only-bullish-or-crisis-periods]
Use time-stratified windows.
### Assuming governance addresses equal people [#assuming-governance-addresses-equal-people]
Analyze delegates and entities where possible, with attribution limits disclosed.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Private groups and deleted content cannot be fully observed.
* Language barriers can bias English-only sampling.
* Researchers can misclassify sarcasm, culture, and technical complexity.
* A small expert community can be healthier than a large general one.
* Incentives can bootstrap real contributors; incentive-linked does not mean fake.
* Pseudonymity can protect users and does not imply bad faith.
* Platform enforcement and outages can distort activity.
Do not produce a single “community score” without retaining the underlying sample.
## Using BlockMind for community research [#using-blockmind-for-community-research]
A BlockMind agent can visit public channels and governance pages with its [browser capability](/docs/capabilities/browser), extract a defined sample, cross-check material claims with current sources, and save the matrix in your Notebook. It cannot reliably determine whether a person is a bot or infer private motives from public text.
Ask:
> “Evaluate \[project] across four time windows. Sample substantive support and critical questions, verify material claims, trace proposals and contributions to outcomes, and flag coordinated-looking patterns without labeling individuals. Show source links, sample rules, counterevidence, and unknowns.”
Verify the sample manually and follow [How to Verify AI Crypto Analysis](/blog/how-to-verify-ai-crypto-analysis).
## The Bottom Line [#the-bottom-line]
A crypto community is valuable when it produces accurate information, resolved user problems, accountable governance, useful contributions, and visible correction under pressure. Audience size is only a reach metric.
Map official channels, sample comparable periods, measure support and output, test critical tolerance, analyze governance, and separate durable participation from incentives and spam. The result should explain how the community works—not how loud it is.
## Sources [#sources]
* [CFTC — Beware Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
* [Discord Safety — How Discord Is Fighting Spam](https://discord.com/safety/how-discord-is-fighting-spam)
* [Discord Safety — Tips Against Spam and Hacking](https://discord.com/safety/360044104071-tips-against-spam-and-hacking)
* [Discord Safety — Identity and Authenticity Policy](https://discord.com/safety/identity-authenticity-policy-explainer-march-2024)
* [GitHub Docs — Default Community Health Files](https://docs.github.com/en/communities/setting-up-your-project-for-healthy-contributions/creating-a-default-community-health-file)
* [GitHub Docs — Adding a Code of Conduct](https://docs.github.com/en/communities/setting-up-your-project-for-healthy-contributions/adding-a-code-of-conduct-to-your-project)
---
# 12 Crypto Due Diligence Mistakes—and the Controls That Prevent Them
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: informational
Primary keyword: crypto due diligence mistakes
Keywords: crypto due diligence mistakes, crypto research mistakes, DYOR mistakes, crypto due diligence red flags, cryptocurrency research checklist
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Most crypto due diligence failures are process failures, not missing indicators. Researchers begin with a conclusion, accept issuer claims as evidence, mix incompatible definitions, ignore executable liquidity, overread transfers and audits, and never set an update trigger. Prevent them with three controls for every material claim: a rule before research, a test during research, and a correction path when evidence changes.
***
## What are the most common crypto due diligence mistakes? [#what-are-the-most-common-crypto-due-diligence-mistakes]
The most damaging mistakes fall into four groups:
* **Framing failures:** researching the wrong asset, horizon, or question
* **Evidence failures:** trusting secondary, stale, or scope-mismatched sources
* **Inference failures:** turning correlation, transfers, audits, or metrics into claims they cannot support
* **Process failures:** skipping contradiction, decision records, and monitoring
This guide focuses on how research breaks. For the full evidence categories, use [What to Check Before Buying Crypto](/blog/what-to-check-before-buying-crypto). For the exhaustive sequence, use the [12-step manual DYOR checklist](/blog/manual-dyor-checklist-12-steps). The two are checklists; this article is a failure-mode pre-mortem.
**Scope and assumptions:** These controls are for public cryptoasset and protocol research. They do not determine legal status, suitability, taxes, or personal risk capacity. Some evidence may remain private or unavailable; mark it unknown rather than filling gaps with confidence.
*This is research, not financial advice. BlockMind’s agent will not tell you what to buy or sell and cannot trade, withdraw, or move funds.*
## The prevent-detect-correct model [#the-prevent-detect-correct-model]
For each failure mode, design three controls:
1. **Prevent:** A rule that reduces the chance of the mistake
2. **Detect:** A check that shows whether it happened
3. **Correct:** A defined action to repair the research record
Example:
| Failure | Prevent | Detect | Correct |
| ------------------- | ---------------------------------- | ------------------------------------------- | --------------------------------------------------- |
| Stale supply figure | Require source date and definition | Compare with current contract/provider data | Recalculate every dependent ratio and note revision |
This is stronger than “be careful” because it creates an observable process.
## Mistake 1: starting with “Should I buy?” [#mistake-1-starting-with-should-i-buy]
That question invites a verdict before the evidence and collapses research, risk, and personal decision-making into one answer.
Use a falsifiable research question instead:
> “Does the current evidence support the project’s claim that protocol usage creates durable demand for this token over a 12-month horizon?”
**Prevent:** Require asset, horizon, claim, and comparison set in the research brief.
**Detect:** If the note contains “bullish” or “bearish” before it defines the claim, restart.
**Correct:** Rewrite the conclusion as observations, interpretation, counterevidence, and invalidation.
The CFTC advises buyers to understand the rights attached to a token and the underlying factors that could affect its value, while warning against guaranteed-value claims ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)). That is a due-diligence question, not a direction call.
## Mistake 2: failing to verify the exact asset [#mistake-2-failing-to-verify-the-exact-asset]
Ticker collisions, wrapped assets, bridges, migrations, test tokens, and fraudulent copies make “research TOKEN” unsafe as an identifier.
**Prevent:** Start with chain, contract or native-asset identifier, official domain, and canonical repository.
**Detect:** Compare at least two official routes to the contract: current docs and explorer link.
**Correct:** Remove evidence attached to the wrong asset and rerun supply, holder, liquidity, and audit checks.
Never copy a contract from a social reply or unsolicited message.
## Mistake 3: treating the project as an independent source [#mistake-3-treating-the-project-as-an-independent-source]
The project is the primary source for what it claims, publishes, and controls. It is not independent confirmation that those claims are true.
**Prevent:** Classify each source as issuer, counterparty, independent primary data, regulator, auditor, analytics provider, or commentary.
**Detect:** Build a claim ledger and count material claims supported only by the issuer.
**Correct:** Verify partnerships through the counterparty, deployments on-chain, releases in canonical repositories, and legal or regulatory claims through the relevant authority.
Use the claim ledger in [How to Read a Crypto Whitepaper](/blog/how-to-read-crypto-whitepaper).
## Mistake 4: using citations that do not support the sentence [#mistake-4-using-citations-that-do-not-support-the-sentence]
A real link can still be irrelevant, stale, or narrower than the claim.
**Prevent:** Place evidence beside the exact claim and record the supporting passage, table, transaction, or field.
**Detect:** Open every decisive citation and ask: same asset, date, unit, definition, and scope?
**Correct:** Narrow the sentence or replace the source. If no source supports it, mark it unverified.
This is especially important for AI-generated research. Follow the [AI verification protocol](/blog/how-to-verify-ai-crypto-analysis) rather than trusting citation formatting.
## Mistake 5: mixing supply and valuation definitions [#mistake-5-mixing-supply-and-valuation-definitions]
Circulating, unlocked, total, and maximum supply are not interchangeable. Provider definitions can differ.
**Prevent:** Put each supply measure, definition, source, and timestamp in a table before calculating valuation.
**Detect:** Recalculate market cap and FDV from the displayed price and supplies; investigate discrepancies.
**Correct:** Use one timestamp and methodology, then update every ratio and comparison.
The [Market Cap vs FDV worksheet](/blog/market-cap-vs-fdv-crypto) shows why a dilution ratio needs an unlock timeline and denominator confidence.
## Mistake 6: assuming every unlock becomes selling [#mistake-6-assuming-every-unlock-becomes-selling]
Vesting, claimability, transferability, circulation, exchange transfer, and sale are separate states.
**Prevent:** Track the state chain for each allocation.
**Detect:** Compare the schedule with vesting-contract state and recipient wallets.
**Correct:** Replace “selling pressure” with the highest verified claim, such as “a tranche equal to 8% of reported circulation becomes claimable.”
Run the complete [token unlock verification protocol](/blog/token-unlocks-vesting-guide).
## Mistake 7: calling a project liquid from market cap or volume [#mistake-7-calling-a-project-liquid-from-market-cap-or-volume]
Neither metric tells you what a relevant order can execute.
**Prevent:** Choose three decision-relevant sizes before looking at venues.
**Detect:** Measure spread, cumulative depth, and simulated impact across credible CEX and DEX routes.
**Correct:** Scope the conclusion by size, time, direction, venue, and stress condition.
FINRA notes that cryptoassets can be less liquid than traditional instruments and that illiquidity can amplify volatility ([FINRA](https://www.finra.org/investors/investing/investment-products/crypto-assets/risks)). Use the reproducible [crypto liquidity test](/blog/crypto-liquidity-analysis) rather than a volume rank.
## Mistake 8: treating “audited” as “safe” [#mistake-8-treating-audited-as-safe]
A report covers specified code and assumptions during a specified engagement. The current deployment can differ.
**Prevent:** Require the auditor-hosted final report, commit, files, exclusions, and dates.
**Detect:** Map current proxy and implementation addresses to the reviewed revision; inspect post-audit changes and unresolved findings.
**Correct:** State what remains covered and what is unverified.
The SEC has separately warned that crypto “proof of reserves” reviews are not equivalent to financial-statement audits under applicable audit standards ([Investor.gov](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins/investors)). Do not confuse a smart-contract audit, reserve attestation, penetration test, and financial audit. They answer different questions.
Use [How to Read a Crypto Audit Report](/blog/how-to-read-crypto-audit-report).
## Mistake 9: reading on-chain transfers as intent [#mistake-9-reading-on-chain-transfers-as-intent]
An exchange inflow is not a sale; an outflow is not necessarily accumulation; one address is not always one entity.
**Prevent:** Use a transfer-to-intent evidence ladder.
**Detect:** Grade labels, remove internal exchange and bridge flows, decode contracts, and inspect entity-level balance changes.
**Correct:** Narrow the claim and preserve the transaction hash.
The full protocol is in [Whale Wallets and Exchange Flows](/blog/whale-wallets-exchange-flows).
## Mistake 10: mistaking attention for adoption [#mistake-10-mistaking-attention-for-adoption]
Followers, messages, searches, and price chatter can grow without users, retention, fees, or contributors.
**Prevent:** Separate reach, engagement, contribution, product usage, and economic output.
**Detect:** Sample support resolution, proposal execution, release-linked contributions, and retention outside incentive windows.
**Correct:** Remove audience metrics from adoption claims unless a reproducible link exists.
The CFTC warns that social channels can coordinate pump-and-dump hype in thin markets ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)). Use the [crypto community substance matrix](/blog/crypto-community-analysis) to measure useful outcomes.
## Mistake 11: collecting only confirming evidence [#mistake-11-collecting-only-confirming-evidence]
More sources do not help if they all repeat the same issuer announcement.
**Prevent:** Before research, write one claim that would support the thesis and one that would disprove it.
**Detect:** Require a “strongest counterevidence” section before the conclusion.
**Correct:** Search primary sources for the disconfirming case, compare peers, and lower confidence if decisive evidence remains unavailable.
Use a contradiction budget:
```text
For every material conclusion:
- strongest supporting evidence
- strongest contradictory evidence
- unresolved evidence gap
- event that would change the conclusion
```
Do not turn the count into a probability.
## Mistake 12: treating research as finished [#mistake-12-treating-research-as-finished]
Crypto evidence decays: contracts upgrade, supply unlocks, maintainers leave, liquidity moves, governance changes, and incidents occur.
**Prevent:** Every conclusion needs a review date and event-based triggers.
**Detect:** At review, test whether source links, deployment addresses, supply, admin roles, liquidity, and thesis conditions changed.
**Correct:** Append a dated change log; do not silently overwrite the original reasoning.
The best workflow preserves what you believed, why, and which new evidence changed it. Copy the [Crypto Research Workflow Template](/blog/crypto-research-workflow-template) for that record.
## Worked pre-mortem: how a “thorough” report fails [#worked-pre-mortem-how-a-thorough-report-fails]
**This example is fictional. It is not a real token assessment or recommendation.**
A 30-page report on “Beacon Token” concludes the project is strong because:
* It has 300 GitHub commits this quarter
* A $900 million market cap and $80 million daily volume
* An audit badge
* A 250,000-member community
* A large exchange outflow
The pre-mortem asks: “Assume this conclusion proves unreliable in 60 days. Which process failures caused it?”
Verification finds:
1. Two unrelated tokens share the ticker; one data source used the wrong contract.
2. Most commits belong to a generated-data repository.
3. Ninety percent of reported volume is on one shallow venue.
4. The audit covers an old implementation and excludes the oracle.
5. Community growth came during a giveaway.
6. The exchange outflow was an internal cold-wallet rotation.
7. A 14% circulating-supply investor cliff was absent from the report.
The report was long but not controlled. Length did not prevent identity, scope, definition, and inference failures.
## The due diligence pre-mortem checklist [#the-due-diligence-pre-mortem-checklist]
Before finalizing, ask:
* Did I verify the exact chain and contract?
* Did I distinguish issuer claims from independent evidence?
* Does every decisive citation support the exact claim?
* Are units, dates, and definitions aligned?
* Did I recalculate material formulas?
* Did I separate unlock from sale and transfer from intent?
* Did I measure liquidity at relevant sizes?
* Does the audit cover the current deployment?
* Did I separate attention from adoption?
* What is the strongest counterevidence?
* Which evidence is unavailable?
* What dated or event-based trigger reopens the research?
If any answer is missing, the conclusion is not ready.
## Limitations and counterevidence [#limitations-and-counterevidence]
* More process cannot eliminate unknowns or future shocks.
* Small projects can lack mature disclosures without being fraudulent.
* Private evidence may be legitimate but not independently verifiable.
* A control can create false comfort if performed mechanically.
* Primary sources can be wrong, self-interested, or later corrected.
* Research quality does not determine market timing or price behavior.
The goal is not to make uncertainty disappear. It is to stop disguising uncertainty as evidence.
## How BlockMind can help without owning the judgment [#how-blockmind-can-help-without-owning-the-judgment]
A BlockMind agent can collect public sources, compare claims, inspect current pages with its [browser capability](/docs/capabilities/browser), review on-chain evidence, and save a dated evidence ledger in your Notebook. It can also monitor named conditions after the first report.
Ask it to run the failure-mode checks:
> “Audit this research note for asset-identity, source-independence, definition, math, audit-scope, liquidity, attribution, confirmation-bias, and freshness failures. Link the evidence, state the strongest contradiction, and mark missing proof as unknown.”
You remain responsible for the interpretation and decision.
## The Bottom Line [#the-bottom-line]
The biggest crypto due diligence mistake is confusing completed activity with reliable research. A long report, many links, or a dozen checked boxes can still rest on the wrong asset, stale definitions, scope gaps, and unsupported inference.
Use prevent, detect, and correct controls. Verify identity, align definitions, trace evidence, test liquidity and deployment, preserve counterevidence, and schedule the next review. Diligence is a maintained process, not a document you finish once.
## Sources [#sources]
* [CFTC — Use Caution When Buying Digital Coins or Tokens](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)
* [CFTC — Beware Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
* [FINRA — Crypto Asset Risks](https://www.finra.org/investors/investing/investment-products/crypto-assets/risks)
* [Investor.gov — Crypto Proof-of-Reserves Reports Are Not Financial Statement Audits](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins/investors)
* [FTC Consumer Advice — What to Know About Cryptocurrency and Scams](https://consumer.ftc.gov/articles/what-know-about-cryptocurrency-scams)
---
# Crypto GitHub Developer Activity: How to Separate Shipping from Commit Theater
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: analyze crypto GitHub activity
Keywords: crypto GitHub developer activity, crypto developer activity, GitHub commits crypto projects, analyze crypto GitHub, open source crypto due diligence
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Raw commit count is a poor measure of crypto development. First map the repositories that actually power the product. Then inspect what changed, whether changes were reviewed and tested, whether releases reach a live deployment, how concentrated maintenance is, and whether security practices match the project’s risk. Developer activity is evidence of work—not proof of adoption, safety, decentralization, or token value.
***
## How do you analyze crypto GitHub developer activity? [#how-do-you-analyze-crypto-github-developer-activity]
Use a six-part evidence chain:
1. **Repository identity:** Is this the canonical codebase for the product?
2. **Change substance:** Do recent changes affect code, tests, docs, generated files, or only automation?
3. **Review and quality:** Are pull requests reviewed, tested, and explained?
4. **Release linkage:** Do commits produce tagged releases and verifiable deployments?
5. **Contributor resilience:** Is maintenance spread across accountable contributors?
6. **Security posture:** Are dependencies, permissions, vulnerabilities, and disclosure handled openly?
Do not start with a leaderboard. A repository can manufacture commits by splitting changes, updating generated files, merging bots, or moving work across branches. Another project can make fewer commits because it is stable or develops critical components in a different repository.
**Scope and assumptions:** This method works best for projects that claim meaningful open-source development. Private repositories, zero-knowledge circuits, hardware, mobile apps, off-chain services, and vendor dependencies may limit visibility. Absence from GitHub is not proof that no work exists; it is a disclosure limitation.
*This is research, not financial advice. BlockMind’s agent does not recommend buys or sells and cannot trade or move funds.*
Repository evidence belongs inside the broader [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto), alongside team, tokenomics, audit, liquidity, holder, and market checks.
## Why contribution graphs and commit counts mislead [#why-contribution-graphs-and-commit-counts-mislead]
GitHub’s own documentation lists important constraints:
* Profile contribution graphs count commits only under specified conditions, including commits on a repository’s default or `gh-pages` branch and an associated author email ([GitHub profile contributions](https://docs.github.com/en/account-and-profile/reference/profile-contributions-reference)).
* Repository contributor graphs show only the top 100 contributors, exclude merge and empty commits, and can omit activity outside the default branch ([GitHub contributors](https://docs.github.com/en/repositories/viewing-activity-and-data-for-your-repository/viewing-a-projects-contributors)).
* Repository-statistics endpoints exclude merge commits; some endpoints are unavailable for repositories with 10,000 or more commits, and statistics are cached based on the default-branch SHA ([GitHub REST API](https://docs.github.com/en/rest/metrics/statistics)).
Therefore “developer count” and “weekly commits” depend on platform rules, branch structure, repository selection, bots, and identity matching. Use them as observations with definitions, not universal productivity measures.
## Step 1: build the canonical repository map [#step-1-build-the-canonical-repository-map]
Start from official current documentation, not GitHub search results.
Map components:
| Product component | Canonical repository | Live artifact | License | Current release |
| ------------------ | -------------------- | ---------------------------- | ------- | --------------- |
| Core protocol/node | | Network/client version | | |
| Smart contracts | | Chain + addresses | | |
| SDK/API | | Package registry | | |
| Front end | | Current app | | |
| Governance | | Proposal/execution contracts | | |
| Docs/specification | | Current docs | | |
Confirm ownership:
* Does the official site link the organization?
* Does the repository link back to the official domain?
* Are releases signed or checksummed?
* Do package names and deployment docs match?
* Was the code moved, archived, forked, or replaced?
Avoid counting tutorial, fork, mirror, testnet, archived, and generated repositories as independent product development unless that is the explicit question.
## Step 2: classify change substance [#step-2-classify-change-substance]
Sample a fixed window, such as the last 90 days and the last 20 merged pull requests. Classify each material change:
| Class | Examples | What it may indicate |
| ---------------------- | ------------------------------------------- | ------------------------------------ |
| Protocol/product | New behavior, performance, integrations | Product evolution |
| Security/correctness | Bug fix, invariant, access control | Risk reduction |
| Test/verification | Unit, integration, fuzz, formal specs | Quality investment |
| Dependency/maintenance | Upgrades, build chain, CI | Maintenance health |
| Documentation | Specs, migration guides, API docs | Usability/transparency |
| Generated/vendor | Lockfiles, generated code, vendored updates | May be mechanical |
| Bot/noise | Formatting churn, automated version bumps | Activity with limited product signal |
Count is less important than the distribution and explanation.
For a sample, record:
* Pull request and issue link
* Problem statement
* Files and components changed
* Tests added or changed
* Reviewers
* Release containing the change
* User- or protocol-visible result
A 10,000-line generated diff can be less substantive than a five-line consensus fix.
## Step 3: inspect review, testing, and issue handling [#step-3-inspect-review-testing-and-issue-handling]
Ask:
* Are changes proposed through pull requests?
* Are approvals from someone other than the author visible?
* Do continuous-integration checks run?
* Are tests changed with behavior?
* Do maintainers explain rejected changes?
* Are security reports routed privately?
* Are breaking changes documented?
OpenSSF Scorecard treats code review, CI tests, branch protection, security policies, dependency updates, pinned dependencies, and known vulnerabilities as separate security-health heuristics. Its maintainers also stress that Scorecard is not definitive and can produce false positives or false negatives ([OpenSSF Scorecard](https://github.com/ossf/scorecard)). Use individual checks as prompts, not the aggregate score as a safety grade.
## Step 4: connect commits to releases and deployments [#step-4-connect-commits-to-releases-and-deployments]
This is the missing link in most “developer activity” charts.
For each material change, verify:
```text
Issue/specification
→ reviewed pull request
→ commit on canonical branch
→ tagged release or package
→ deployment or network activation
→ current documentation
```
Record where the chain breaks.
For smart contracts:
* Map source revision to verified deployed bytecode
* Resolve proxies and current implementation addresses
* Check upgrade events after the latest review
* Compare with [the audit report scope](/blog/how-to-read-crypto-audit-report)
For node software:
* Check release notes, signed artifacts, supported versions, and network upgrade status
For SDKs:
* Check registry publication, package provenance, documentation, and whether dependent applications adopted the release
Work that never ships may still be valuable research, but it should not be described as a live product improvement.
## Step 5: measure contributor resilience [#step-5-measure-contributor-resilience]
Instead of counting every contributor equally, inspect maintenance concentration.
### Contribution concentration [#contribution-concentration]
For a fixed window:
```text
Top-contributor share = commits or merged changes by top contributor / total sampled changes
Top-three share = changes by top 3 contributors / total sampled changes
```
Use merged pull requests or reviewed change sets when possible, not lines changed.
Then ask:
* How many people can review and release critical components?
* Are maintainers from one employer or several organizations?
* Are external contributions accepted and meaningfully reviewed?
* Is succession documented?
* Are keys, releases, and package publication controlled by one person?
High concentration can be efficient in an early project and dangerous for operational resilience. Low concentration can represent a healthy community or fragmented accountability. Context matters.
### Identity caveats [#identity-caveats]
One developer can use several emails; bots can appear as contributors; rebases can change attribution; squashed pull requests can compress many commits into one. Disclose the counting unit and manually inspect the top contributors.
## Step 6: evaluate maintenance and security signals [#step-6-evaluate-maintenance-and-security-signals]
Check the repository at the component level:
* License and copyright
* `SECURITY.md` and private-reporting route
* Supported versions
* Dependency update process
* Reproducible build instructions
* Release signing and provenance
* Branch or ruleset protection
* Required review
* CI and security scanning
* Open known vulnerabilities and advisories
* Incident postmortems
* Bus-factor and key-management disclosure
GitHub’s community-health files include contributing, governance, security, support, and code-of-conduct documents that make project processes easier to inspect ([GitHub Docs](https://docs.github.com/en/communities/setting-up-your-project-for-healthy-contributions/creating-a-default-community-health-file)). Their presence is evidence of documented process; their absence is not proof no process exists.
## The developer-activity evidence matrix [#the-developer-activity-evidence-matrix]
| Dimension | Stronger evidence | Counterevidence | Confidence |
| ----------------- | ------------------------------------------ | --------------------------- | ---------- |
| Canonical mapping | Official links and live-artifact match | Unclear or abandoned repos | |
| Change substance | Product/security/test changes | Generated or bot churn | |
| Review | Independent review and passing checks | Direct unreviewed pushes | |
| Releases | Regular explained releases | Commits never shipped | |
| Deployment | Verifiable version/address match | Live code unknown | |
| Contributors | Multiple accountable maintainers | Single-person critical path | |
| Security | Disclosure, advisories, dependency hygiene | No reporting path | |
| Documentation | Current specs and migration notes | Whitepaper-only claims | |
Do not average this into a token score. Preserve which components are strong and which remain unknown.
## Worked hypothetical: “Atlas Chain” [#worked-hypothetical-atlas-chain]
**Atlas Chain is fictional. The observations are illustrative, not a real project assessment.**
An analytics site reports 1,400 Atlas commits in the last 90 days and calls it the “most active chain.”
The repository map finds six public repos. Review shows:
* 980 commits are automated snapshots in a generated-data repository.
* 220 are dependency and formatting updates.
* 110 are documentation changes.
* 90 touch node or contract code.
* Of those 90, 46 belong to one unreleased experimental branch.
* The remaining 44 map to 12 merged pull requests, each reviewed and tested.
* Two releases include nine of those pull requests; the latest network version includes seven.
* Three maintainers authored 88% of sampled merged changes, and one person controls the current package release process.
* A security policy and bug bounty exist, but the bounty excludes the bridge contract holding most user-deposited value.
Responsible conclusion:
> Atlas shows real reviewed development, but the headline commit count is dominated by generated activity. Seven material changes reached the current network during the window. Maintenance and release authority are concentrated, and the highest-value bridge sits outside the stated bounty scope.
That is more decision-useful than “1,400 commits.”
## Questions raw activity cannot answer [#questions-raw-activity-cannot-answer]
Developer evidence does not prove:
* Users want the product
* Fees or revenue are sustainable
* Tokenholders capture product value
* The protocol is legally compliant
* Smart contracts are safe
* Governance is decentralized
* The team will continue contributing
* The token is fairly valued
Combine repository analysis with [whitepaper verification](/blog/how-to-read-crypto-whitepaper), [community analysis](/blog/crypto-community-analysis), market and protocol fundamentals, and current deployments.
## Common GitHub analysis mistakes [#common-github-analysis-mistakes]
### Ranking by commits without repository mapping [#ranking-by-commits-without-repository-mapping]
You may count forks, mirrors, generated data, experiments, and obsolete code.
### Treating lines added as progress [#treating-lines-added-as-progress]
Deletions and simplification can improve software. Generated files can inflate additions.
### Counting all contributors as maintainers [#counting-all-contributors-as-maintainers]
One typo fix and release authority are not equivalent roles.
### Ignoring default-branch and graph rules [#ignoring-default-branch-and-graph-rules]
GitHub statistics have documented exclusions. Record the query and branch.
### Assuming open source equals deployed source [#assuming-open-source-equals-deployed-source]
Verify bytecode, packages, releases, and network versions.
### Assuming more change is always safer [#assuming-more-change-is-always-safer]
Frequent change increases review and upgrade burden. Mature critical code may change carefully.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Important work may occur in private repositories or non-GitHub forges.
* Repository history can be rewritten or imported.
* Automated classifications can mislabel changes.
* Commit identity does not necessarily map to legal identity or employer.
* Public review comments do not reveal every offline review.
* A quiet repository can be abandoned or stable; inspect releases, issues, and support policy.
* Security heuristics are context-dependent and can miss controls implemented elsewhere.
State what you could not see.
## Using BlockMind to review developer activity [#using-blockmind-to-review-developer-activity]
A BlockMind agent can locate current official repository links, inspect release pages and pull requests with its [browser capability](/docs/capabilities/browser), classify a defined sample, and save the repository map in your Notebook.
Ask for reproducibility:
> “Map the official repositories for \[project]. Over the last 90 days, classify the 20 most recent merged pull requests by product, security, test, maintenance, docs, or generated work. Link each material change to a release and live deployment where possible. Disclose GitHub API exclusions, bots, branches, and unresolved identity or deployment gaps.”
Then verify the decisive sample manually. AI can misclassify diffs and repository ownership; follow [How to Verify AI Crypto Analysis](/blog/how-to-verify-ai-crypto-analysis).
## The Bottom Line [#the-bottom-line]
Crypto GitHub activity becomes useful only when it connects a canonical repository to substantive reviewed changes, releases, and the live product. Count less; trace more.
Map the codebase, sample change quality, inspect review and tests, verify deployment, measure maintenance concentration, and record security practices. The result should describe evidence of shipping and resilience—not turn commits into a price signal.
## Sources [#sources]
* [GitHub Docs — Profile Contributions Reference](https://docs.github.com/en/account-and-profile/reference/profile-contributions-reference)
* [GitHub Docs — Viewing a Project’s Contributors](https://docs.github.com/en/repositories/viewing-activity-and-data-for-your-repository/viewing-a-projects-contributors)
* [GitHub Docs — REST API Repository Statistics](https://docs.github.com/en/rest/metrics/statistics)
* [GitHub Docs — Default Community Health Files](https://docs.github.com/en/communities/setting-up-your-project-for-healthy-contributions/creating-a-default-community-health-file)
* [OpenSSF Scorecard — Goals, Non-Goals, and Checks](https://github.com/ossf/scorecard)
* [OpenSSF Scorecard — Detailed Check Documentation](https://github.com/ossf/scorecard/blob/main/docs/checks.md)
---
# How to Check Crypto Holder Distribution Without Misreading It
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: crypto holder distribution
Keywords: crypto holder distribution, token holder distribution, how to check token holders, crypto whale concentration, top token holders
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Do not judge token concentration from the explorer’s top-holder table alone. First verify the contract and supply denominator, then classify exchanges, bridges, liquidity pools, treasuries, vesting contracts, burn addresses, and likely independent wallets. Report both raw and adjusted concentration, trace changes over time, and state which labels are uncertain. One address is not necessarily one person, and one person can control many addresses.
Holder distribution answers a useful risk question: how much influence over liquid supply may be concentrated among a small number of economic actors? A raw address ranking does not answer it by itself.
It is one evidence category inside the full [what to check before buying crypto](/blog/what-to-check-before-buying-crypto) workflow, not a standalone safety verdict.
An exchange omnibus wallet may represent thousands of customers. A liquidity-pool contract holds tokens on behalf of market participants. A vesting contract may be locked but economically allocated to insiders. A bridge escrow address may mirror supply on another chain. Treating all of them as comparable “whales” produces a precise-looking mistake.
## What token holder distribution measures [#what-token-holder-distribution-measures]
For a standard ERC-20 token, the contract exposes `totalSupply()` and `balanceOf(address)`. The standard defines those functions, while transfer events allow indexers to reconstruct balance changes ([ERC-20 specification](https://ercs.ethereum.org/ERCS/erc-20)). Explorers and APIs turn those records into a ranked holder list; Etherscan, for example, documents an endpoint that returns current ERC-20 holder addresses and quantities ([Etherscan](https://docs.etherscan.io/api-reference/endpoint/tokenholderlist)).
The raw concentration formula is:
```text
Top-N concentration = sum(balance of top N addresses) / chosen supply denominator
```
Every term needs a definition. “Top N” before or after exclusions? Balance at which block? Total, circulating, or liquid supply? Are mirrored bridge balances double-counted? A result without those choices cannot be compared responsibly.
## The nine-step holder-distribution workflow [#the-nine-step-holder-distribution-workflow]
### 1. Verify the canonical contract [#1-verify-the-canonical-contract]
Resolve the official contract from project documentation and cross-check it with the chain’s explorer and a reputable registry. Confirm chain, decimals, token standard, proxy or upgrade pattern, and whether the asset has native, wrapped, or bridged versions.
Ticker search is not enough. Scam tokens routinely reuse names and symbols.
### 2. Choose the observation block and timestamp [#2-choose-the-observation-block-and-timestamp]
Current holder lists change. Record a block height and UTC timestamp so another researcher can reproduce the snapshot. If comparing two dates, use the same classification and denominator method.
### 3. Reconcile the supply denominator [#3-reconcile-the-supply-denominator]
Collect:
* on-chain total supply;
* reported circulating supply;
* maximum supply, if defined;
* burned or irrecoverable supply;
* locked and vesting allocations;
* bridge escrow and mirrored supply;
* treasury and protocol-owned balances.
Do not assume “circulating supply” is directly encoded in a token contract. It is often a provider or issuer calculation based on exclusions.
Report at least two views when material:
1. **Share of on-chain total supply** for reproducibility.
2. **Share of estimated liquid or circulating supply** with the exclusion method stated.
### 4. Classify the top addresses [#4-classify-the-top-addresses]
Use explorer labels, contract code, project disclosures, transaction counterparties, and specialist entity data. Apply categories rather than simply deleting addresses:
| Category | Interpretation question |
| ------------------------ | -------------------------------------------------------------- |
| Burn/null | Is the balance provably inaccessible? |
| Exchange/custodian | Does one address represent many beneficial owners? |
| Bridge escrow | Is supply represented on another chain? |
| Liquidity pool | What share belongs to LPs, and how concentrated are LP tokens? |
| Vesting contract | Who is the beneficiary, and when can tokens unlock? |
| Team/treasury/foundation | Who controls keys and governance over transfers? |
| Staking/vault contract | Are depositors independently redeemable? |
| Protocol-owned liquidity | Is control centralized despite an AMM label? |
| Unknown EOA/entity | Could related addresses share one controller? |
Address labels add interpretation, not certainty. Nansen’s address-label endpoint, for example, exposes entity, behavioral, DeFi, and centralized-exchange labels ([Nansen](https://docs.nansen.ai/api/profiler/address-labels)). Preserve “unknown” when evidence is weak.
The broader distinction between raw transfers, decoded actions, and entity inference is explained in [what on-chain analysis is](/blog/what-is-on-chain-analysis-crypto).
### 5. Calculate raw and adjusted concentration [#5-calculate-raw-and-adjusted-concentration]
Report several views rather than one magic threshold:
* top 1, 5, 10, and 20 addresses as a share of total supply;
* the same after excluding burn and clearly identified infrastructure;
* insider/treasury/vesting share separately;
* unknown top-holder share;
* liquid-pool and exchange custody share;
* concentration by labeled entity where clustering is defensible.
An adjusted formula might be:
```text
Adjusted top-N concentration =
sum(top N independent/insider balances) /
(total supply - provably burned - non-circulating bridge duplication)
```
Do not hide the raw number. Adjustments improve interpretation but add subjective choices.
### 6. Inspect control, not just ownership [#6-inspect-control-not-just-ownership]
For contracts and organizational wallets, ask:
* Is it a multisig, and what is the signing threshold?
* Are signers independent or controlled by one organization?
* Is there a timelock?
* Can an admin upgrade the contract or change transfer restrictions?
* Can tokens be minted, frozen, paused, or clawed back?
* Are vesting beneficiaries and schedules public?
* Can treasury governance be changed quickly by the same holders?
Distribution can look broad while effective control remains narrow.
### 7. Analyze changes, not only the snapshot [#7-analyze-changes-not-only-the-snapshot]
A trend often matters more than a single rank. Compare consistent snapshots and trace:
* team or vesting transfers;
* treasury outflows;
* deposits to and withdrawals from labeled exchanges;
* accumulation across clusters of addresses;
* liquidity migration;
* bridge movements;
* new mint or burn events.
A transfer to an exchange increases the possibility of sale; it does not prove a sale. A withdrawal may indicate self-custody, internal exchange reorganization, or another purpose. State the observation and inference separately.
### 8. Connect concentration to liquidity [#8-connect-concentration-to-liquidity]
A holder with 2% of supply can be more consequential than a holder with 10% if the tradable float and pool liquidity are tiny. Compare potential sell size with:
* pool reserves and active order-book depth;
* typical daily volume after suspected wash activity;
* unlock schedule;
* market-maker arrangements if disclosed;
* distribution across venues.
Holder risk is a relationship between control, float, liquidity, and behavior—not a universal percentage cutoff.
### 9. Write a reproducible conclusion [#9-write-a-reproducible-conclusion]
Use a format like:
> At block X, the top 10 addresses held A% of on-chain total supply. After excluding the burn address, bridge escrow, two exchange custody wallets, and a liquidity-pool contract, the top 10 remaining addresses held B% of the adjusted denominator. C% remained in unknown top addresses. Team, treasury, and vesting-controlled balances accounted for D%, with the next material unlock on date E according to source F. Labels are heuristic; no sale is inferred from transfers alone.
## How to interpret common patterns [#how-to-interpret-common-patterns]
### One giant exchange wallet [#one-giant-exchange-wallet]
This is custody concentration, not automatically investor concentration. It still creates operational and venue risk, but do not call every underlying customer one whale.
### Many newly funded wallets with similar behavior [#many-newly-funded-wallets-with-similar-behavior]
Nominal distribution may overstate independence. Look for common funding sources, synchronized transfers, identical timing, and shared counterparties. Clustering is probabilistic; do not claim common ownership without evidence.
### A large locked vesting contract [#a-large-locked-vesting-contract]
Locked supply may not create immediate selling pressure, but it represents future economic allocation. Record beneficiary, cliff, cadence, admin powers, and whether the contract is immutable.
### A large liquidity-pool balance [#a-large-liquidity-pool-balance]
The pool address is not one holder. Inspect LP-token or position ownership, removable liquidity, and whether the project controls most liquidity. “Locked liquidity” also needs expiry, locker contract, and beneficiary verification.
### Burn address balances [#burn-address-balances]
Only exclude supply that is provably inaccessible under the relevant chain and contract behavior. A branded “dead” wallet is not enough if someone can control it.
### Cross-chain supply [#cross-chain-supply]
Bridge escrow can look like a huge holder on the origin chain while tokens circulate on a destination chain. Reconcile canonical and wrapped supply to avoid double counting.
## Why address count is not holder count [#why-address-count-is-not-holder-count]
Glassnode’s entity research explains both directions of the problem: one entity can control multiple Bitcoin addresses, while one address—such as an exchange—can hold funds for many users ([Glassnode](https://research.glassnode.com/bitcoin-supply-distribution/)). The same conceptual warning applies when interpreting token holders.
“Number of holders increased” can also reflect dust transfers, airdrops, sybil accounts, or users splitting custody. Combine it with balance distribution and economically meaningful activity.
## Questions to ask an AI analyst [#questions-to-ask-an-ai-analyst]
* “Resolve the official contract and stop if identity is ambiguous.”
* “Show raw top-holder balances at a stated block before applying labels.”
* “Classify each exclusion and cite the evidence.”
* “Calculate raw and adjusted top-10 concentration using both total and estimated liquid supply.”
* “Group only high-confidence related addresses and keep an ungrouped view.”
* “Trace material team, treasury, and vesting transfers without calling them sales unless swap or venue evidence supports that.”
* “List unknown wallets large enough to change the conclusion.”
BlockMind’s [on-chain intelligence](/docs/capabilities/on-chain) can help interpret holder structure and labeled flows. The output still needs the same methodological discipline.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Address labels can be incomplete, stale, or wrong.
* Privacy systems, centralized venues, and off-chain legal arrangements hide beneficial ownership.
* One entity can split balances to appear decentralized.
* An exchange address aggregates unrelated users.
* Supply APIs can disagree about circulating exclusions.
* A concentrated token can be legitimate; a dispersed token can still be coordinated or unsafe.
* Holder distribution says little about product quality by itself.
Do not use arbitrary rules such as “top 10 above 50% is always a scam.” Context can reverse the interpretation. Use concentration as one part of [complete crypto due diligence](/blog/manual-dyor-checklist-12-steps).
## The Bottom Line [#the-bottom-line]
The top-holder page is the beginning of holder analysis, not the conclusion. Verify the contract, freeze the snapshot at a block, reconcile supply, classify addresses, calculate raw and adjusted concentration, examine control and liquidity, and trace changes. Report uncertainty instead of converting every address into a person.
This is research, not financial advice. BlockMind’s agent can analyze on-chain evidence but never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [ERC-20 Token Standard](https://ercs.ethereum.org/ERCS/erc-20)
* [Etherscan — Token Holder List API](https://docs.etherscan.io/api-reference/endpoint/tokenholderlist)
* [Nansen API — Address Labels](https://docs.nansen.ai/api/profiler/address-labels)
* [Glassnode Research — Bitcoin Supply Distribution and Entities](https://research.glassnode.com/bitcoin-supply-distribution/)
* [Ethereum.org — Block Explorers](https://ethereum.org/developers/docs/data-and-analytics/block-explorers/)
---
# Crypto Liquidity Analysis: Measure Depth, Slippage, and Exit Capacity
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: crypto liquidity analysis
Keywords: crypto liquidity analysis, crypto market depth, crypto slippage, DEX liquidity analysis, order book depth crypto
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Crypto liquidity is the capacity to transact a relevant size, at a known time, without an unacceptable price change. Measure it with spread, cumulative order-book depth, executable quote impact, DEX active liquidity, and venue concentration. Market cap and reported daily volume are context; neither tells you what your order can execute today.
***
## What is crypto liquidity analysis? [#what-is-crypto-liquidity-analysis]
Crypto liquidity analysis measures how easily an asset can be exchanged at or near a reference price. A useful analysis is specific to:
* Asset and pair
* Buy or sell direction
* Order size
* Venue or route
* Time and volatility regime
* Execution method
“Token X is liquid” is incomplete. “At 14:00 UTC, a $25,000 simulated sell across the two largest credible venues showed 0.6% blended impact, while a $250,000 order exceeded 4%” is reproducible.
**Scope and assumptions:** This is a research and market-quality framework, not execution advice. Do not place trades merely to test liquidity. Use read-only order books, venue quote endpoints, or transaction simulations where possible. Fees, gas, MEV, latency, account tier, routing, and market movement can make actual execution differ from a quote.
*This is research, not financial advice. BlockMind’s agent cannot trade, withdraw, or move funds and will not tell you what to buy or sell.*
Liquidity is one required layer in the broader [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto). It should be checked alongside token rights, supply, security, team evidence, and the reason the token exists.
## Why market cap and 24-hour volume are not liquidity [#why-market-cap-and-24-hour-volume-are-not-liquidity]
Market capitalization is a reference price multiplied by a supply estimate. It does not represent bids available to absorb a sale.
Reported volume is activity over a period. It does not reveal:
* How much depth is available now
* Whether volume is concentrated on one venue
* Whether the venue is accessible or credible
* Whether volume is organic, incentivized, or inflated
* Whether the pair uses a risky quote asset
* How much price impact a given order creates
The CFTC has warned that some virtual-currency cash markets can be unregulated or unsupervised and that participants should understand platform and product risks ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/understand_risks_of_virtual_currency.html)). Therefore, usable liquidity includes venue quality and access—not just displayed size.
## The five measurements that matter [#the-five-measurements-that-matter]
### 1. Bid-ask spread [#1-bid-ask-spread]
For best bid `B` and best ask `A`:
```text
Mid-price = (A + B) / 2
Absolute spread = A − B
Spread % = (A − B) / Mid-price × 100%
```
A tight spread indicates close top-of-book quotes. It does not prove meaningful depth behind them. A tiny order can remove the best quote.
### 2. Cumulative order-book depth [#2-cumulative-order-book-depth]
Sum visible bids and asks within standardized distances from the mid-price, such as 0.5%, 1%, 2%, and 5%.
```text
Bid depth at x% = value of bids between mid and mid × (1 − x)
Ask depth at x% = value of asks between mid and mid × (1 + x)
```
Record both sides. A book can have strong bid depth and weak ask depth, or vice versa.
Displayed depth can be canceled, duplicated across related venues, or affected by hidden orders. Treat it as a snapshot, then repeat the observation.
### 3. Executable price impact [#3-executable-price-impact]
For a simulated order with reference mid-price `Pmid` and average execution price `Pavg`:
```text
Buy impact % = (Pavg − Pmid) / Pmid × 100%
Sell impact % = (Pmid − Pavg) / Pmid × 100%
```
Uniswap defines price impact as the price change caused directly by the trade and explains that pool size affects the result ([Uniswap Labs](https://support.uniswap.org/hc/en-us/articles/8671539602317-What-is-Price-Impact)).
Do not confuse **price impact** with **slippage**. Price impact arises from consuming available liquidity. Slippage is the difference between expected and actual execution and can also reflect price movement while the order is processed. Coinbase notes that a market order can fill at a less favorable price than the most recent trade depending on available volume and prices ([Coinbase](https://help.coinbase.com/en/coinbase/trading-and-funding/advanced-trade/order-management)).
### 4. DEX active liquidity and route quality [#4-dex-active-liquidity-and-route-quality]
In a constant-product pool:
```text
x × y = k
```
Removing one asset and adding the other changes their reserve ratio and therefore the price. Uniswap’s documentation describes this reserve-based price mechanism ([Uniswap Labs](https://support.uniswap.org/hc/en-us/articles/7422670207373-How-are-token-prices-determined)).
For concentrated-liquidity designs, headline TVL can overstate liquidity available near the current price because providers choose ranges. Check:
* Active liquidity near current price
* Pool fee tier
* Route through intermediate assets
* Quote-token quality
* Gas and protocol fees
* Pool and LP concentration
* Whether liquidity can be removed by a small set of wallets
Uniswap explicitly warns that a team acting as primary liquidity provider may be able to remove liquidity if it is not locked ([Uniswap Labs](https://support.uniswap.org/hc/en-us/articles/37113550065549-What-are-the-risks-when-providing-liquidity)). “Locked liquidity” itself still requires contract, duration, beneficiary, and unlock verification.
### 5. Venue concentration and resilience [#5-venue-concentration-and-resilience]
Calculate the share of credible volume and depth on each venue:
```text
Venue concentration = largest venue’s credible volume / total credible volume
```
Also calculate for 1% depth. High concentration introduces venue-specific risk: outage, withdrawal suspension, regional access, delisting, or a quote-asset problem.
Resilience asks whether liquidity returns after a shock. Repeat measurements during normal conditions, volatility, and after material events.
## The standardized liquidity test [#the-standardized-liquidity-test]
Use the same protocol for every asset.
### Step 1: fix the timestamp and pairs [#step-1-fix-the-timestamp-and-pairs]
Record UTC time, asset contract or identifier, pairs, venues, and quote assets. Exclude ambiguous tickers.
### Step 2: select decision-relevant sizes [#step-2-select-decision-relevant-sizes]
Use at least three sizes, for example:
* Small: the minimum meaningful research size
* Medium: a realistic portfolio-change size
* Large: a stress size
Do not publish a universal dollar ladder as a safety threshold. Relevant size depends on the reader and asset.
### Step 3: capture CEX book quality [#step-3-capture-cex-book-quality]
For each credible venue:
* Spread
* Bid and ask depth at 0.5%, 1%, 2%, and 5%
* Estimated average fill at each standardized size
* Pair volume and timestamp
* Deposit/withdrawal status, if material
### Step 4: capture DEX route quality [#step-4-capture-dex-route-quality]
For each route:
* Chain, pool, fee tier, and token contract
* Quoted output at each size
* Price impact
* Gas and protocol fees shown separately
* Active-liquidity range where available
* LP and removable-liquidity concentration
### Step 5: repeat and summarize a range [#step-5-repeat-and-summarize-a-range]
One snapshot can be lucky. Repeat at several times or query a historical depth series. Report median and worst observed result, sample count, and observation window.
## Liquidity scorecard without a fake magic number [#liquidity-scorecard-without-a-fake-magic-number]
| Dimension | Stronger evidence | Warning evidence | Result |
| ----------------- | -------------------------------- | ----------------------- | ------ |
| Spread | Consistently tight across venues | Wide or unstable | |
| Depth | Balanced at relevant size | Shallow or one-sided | |
| Price impact | Low and stable at tested sizes | Nonlinear jump | |
| Venue diversity | Several credible routes | One dominant venue | |
| Quote quality | Deep, reputable quote assets | Fragile/obscure quote | |
| DEX control | Distributed active liquidity | Team/LP concentration | |
| Stress resilience | Depth replenishes | Liquidity disappears | |
| Data confidence | Reproducible methods | Conflicting/opaque data | |
Use labels—adequate, constrained, or fragile—for the specific tested size and time. Preserve the underlying measurements.
## Worked hypothetical: “Granite” looks liquid until size changes [#worked-hypothetical-granite-looks-liquid-until-size-changes]
**Granite and all figures below are fictional. This is not a real market test, backtest, or recommendation.**
Granite shows a $600 million market cap and $40 million reported 24-hour volume.
At 12:00 UTC:
| Test | $5,000 sell | $50,000 sell | $250,000 sell |
| ---------------------- | ----------: | -----------: | ------------: |
| CEX A estimated impact | 0.08% | 0.55% | 3.7% |
| CEX B estimated impact | 0.12% | 0.90% | 6.4% |
| Main DEX route | 0.20% | 1.8% | 11.5% |
Additional observations:
* CEX A holds 72% of credible 1% bid depth.
* Most DEX liquidity is concentrated in a narrow range just above current price.
* The project treasury supplies 43% of active DEX liquidity through two disclosed wallets.
* During a market-wide volatility window, $50,000 impact on CEX A temporarily rises from 0.55% to 2.1%.
Conclusion:
> Granite is liquid for the small tested size under ordinary conditions, constrained at the medium size, and fragile at the stress size. Liquidity is concentrated by venue and LP, and degrades materially during volatility. The headline market cap and volume did not reveal those limits.
The conclusion is scoped. It does not call the asset universally liquid or illiquid.
## How to analyze liquidity around token unlocks [#how-to-analyze-liquidity-around-token-unlocks]
Before a [token unlock](/blog/token-unlocks-vesting-guide), compare the tranche with:
* Current 1% and 2% bid depth
* Standardized sell-quote impact
* Median credible daily volume
* Historical recipient transfers
* Market-maker and treasury liquidity commitments
Do not divide unlock notional by daily volume and call the answer “days to absorb.” Volume includes both sides and repeated turnover, and not all volume is available demand. Use it only as rough context with depth and quote tests.
## Liquidity red flags [#liquidity-red-flags]
* High reported volume with wide spread and shallow depth
* Most activity on an obscure or inaccessible venue
* Different assets share the same ticker
* One quote asset dominates and has its own stability risk
* DEX liquidity is outside the active price range
* A team wallet controls most LP tokens and can remove them
* Depth appears briefly and disappears when approached
* Deposits or withdrawals are suspended on the dominant venue
* Aggregators route through a risky intermediate token
* A small test order causes nonlinear impact
* Liquidity claims have no timestamp, size, pair, or methodology
## Limitations and counterevidence [#limitations-and-counterevidence]
* Order books change continuously and displayed orders may be canceled.
* Simulations do not guarantee execution.
* DEX execution can be affected by MEV, gas, routing, and state changes.
* Venue APIs can omit hidden liquidity.
* OTC liquidity is usually not public.
* A low-liquidity asset can become deeper, and a deep market can disappear during stress.
* Locked LP positions do not remove smart-contract, range, quote-asset, or post-unlock withdrawal risk.
Refresh the test whenever position size, volatility, dominant venue, token supply, or pool ownership changes.
## Using BlockMind for liquidity research [#using-blockmind-for-liquidity-research]
A BlockMind agent can compare current public market and on-chain sources, inspect live venue or pool pages with its [browser capability](/docs/capabilities/browser), and review DEX volume and holder context through [on-chain intelligence](/docs/capabilities/on-chain). It cannot execute the test as a real trade or move funds.
Ask for an auditable output:
> “For \[asset and contract], measure spread, 0.5/1/2/5% bid and ask depth, and simulated price impact at \[three sizes] across named credible venues and the main DEX routes. Timestamp every observation, separate fees from impact, disclose exclusions, and classify liquidity only for those tested sizes.”
Save the results in your Notebook and combine them with the broader [Crypto Market Analysis Guide](/blog/crypto-market-analysis-guide).
## The Bottom Line [#the-bottom-line]
Liquidity is not market cap, volume, or TVL. It is executable capacity at a stated size and time. Measure spread, depth, simulated impact, active DEX liquidity, venue concentration, and resilience under stress.
A good liquidity conclusion is deliberately narrow: adequate for this size under these conditions, constrained beyond that size, and subject to these venue, pool, and data limitations.
## Sources [#sources]
* [Uniswap Labs — What Is Price Impact?](https://support.uniswap.org/hc/en-us/articles/8671539602317-What-is-Price-Impact)
* [Uniswap Labs — How Are Token Prices Determined?](https://support.uniswap.org/hc/en-us/articles/7422670207373-How-are-token-prices-determined)
* [Uniswap Labs — Risks When Providing Liquidity](https://support.uniswap.org/hc/en-us/articles/37113550065549-What-are-the-risks-when-providing-liquidity)
* [Uniswap Developers — How Uniswap AMMs Work](https://developers.uniswap.org/docs/get-started/concepts/how-uniswap-works)
* [Coinbase — Order Management and Slippage](https://help.coinbase.com/en/coinbase/trading-and-funding/advanced-trade/order-management)
* [CFTC — Understand the Risks of Virtual Currency Trading](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/understand_risks_of_virtual_currency.html)
---
# Crypto Market Analysis: A Reproducible Top-Down Guide
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: how to analyze the crypto market
Keywords: crypto market analysis, how to analyze crypto market, crypto fundamental analysis, crypto technical analysis, crypto market indicators
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Good crypto market analysis is not a hunt for one perfect indicator. It is a repeatable process: define the timeframe, classify the market regime, inspect price structure, verify participation, examine leverage and liquidity, then write down the evidence that would disprove your interpretation. A conclusion is strongest when independent layers agree and weakest when one dramatic chart carries the entire case.
***
## What is crypto market analysis? [#what-is-crypto-market-analysis]
Crypto market analysis is the structured interpretation of price, volume, positioning, liquidity, on-chain activity, fundamentals, and catalysts over a stated timeframe. Its purpose is not to predict every move. Its purpose is to turn a noisy market into a testable description of what is happening, why it may be happening, and what evidence would change that description.
This guide uses a **six-layer evidence matrix**:
1. Regime: the broad financial and crypto environment
2. Structure: the direction and shape of price
3. Participation: whether activity confirms the move
4. Leverage: whether derivatives positioning makes the move fragile
5. Liquidity: whether the quoted market can absorb realistic size
6. Catalysts and fundamentals: what could sustain or break the move
That order matters. A token chart can look strong inside a weakening market; reported volume can look large while usable depth is thin; and rising price can coexist with increasingly crowded leverage. Each layer answers a different question.
**Scope and assumptions:** This is a research workflow for liquid cryptoassets and protocols with verifiable public data. It does not provide a valuation model for every token, and it is not a trading system. Data definitions vary by venue and provider, so record the source, timestamp, pair, and timeframe for every observation.
*BlockMind provides research, not financial advice. Your agent will not tell you what to buy or sell and cannot trade, withdraw, or move funds.*
## The six-layer crypto market analysis framework [#the-six-layer-crypto-market-analysis-framework]
### 1. Define the question before opening a chart [#1-define-the-question-before-opening-a-chart]
Write a one-sentence question with an asset, horizon, and decision context:
> “Has Asset A’s four-week advance become broad and durable enough to justify deeper research, or is it a thin, leverage-led move?”
This prevents timeframe leakage. A constructive weekly structure and an overextended hourly chart can both be true. They only conflict if you silently switch horizons.
Record four items:
* **Asset and reference pair:** for example, TOKEN/USD and TOKEN/BTC
* **Research window:** such as 90 days, one year, or a full cycle
* **Decision window:** days, months, or years
* **Comparison set:** Bitcoin, a sector index, and two direct peers
The reference pair changes the claim. TOKEN/USD strength may simply reflect a rising Bitcoin market. TOKEN/BTC strength asks whether the token is outperforming Bitcoin.
### 2. Classify the regime [#2-classify-the-regime]
Regime is the background that affects how much confidence to place in asset-specific signals. Use both macro and crypto-native context.
At minimum, record:
* The direction of major policy rates and the next scheduled policy events
* Broad risk-asset trend, without assuming correlation is permanent
* Bitcoin trend and [Bitcoin dominance](/blog/bitcoin-dominance-explained)
* Breadth: how much of the market is participating
* A sentiment input such as the [Fear & Greed Index](/blog/fear-greed-index-guide), treated as context rather than a trigger
The Federal Reserve explains that changes in its policy-rate target influence other short-term rates and economic activity ([Federal Reserve](https://www.federalreserve.gov/economy-at-a-glance-policy-rate.htm)). That does not produce a mechanical crypto signal. It is a reason to document the rate environment rather than analyze a token in isolation.
Use three regime labels instead of “bull” and “bear”:
| Regime | Evidence pattern | Research implication |
| ----------- | --------------------------------------------------------------------------------- | --------------------------------------------------------- |
| Expansion | Broad participation, improving liquidity, constructive higher-timeframe structure | Asset-specific strength has a friendlier backdrop |
| Transition | Conflicting breadth, leadership changes, volatility shifts | Reduce certainty; look for confirmation and failed breaks |
| Contraction | Weak breadth, deteriorating liquidity, repeated lower highs | Demand stronger asset-specific evidence |
These labels describe evidence; they do not prescribe a position.
### 3. Map price structure without narrating every candle [#3-map-price-structure-without-narrating-every-candle]
Price structure answers: **what direction has the market actually sustained?**
Mark only the features another researcher could reproduce:
* Major swing highs and lows
* Areas repeatedly accepted or rejected
* Trend relative to the chosen comparison asset
* Volatility expansion or compression
* The level that would invalidate the current structure label
Avoid drawing a line first and inventing a story second. A useful note sounds like this:
> “On the weekly horizon, price has made two higher swing lows and closed above the prior range. The structure returns to neutral if a weekly close is accepted back inside that range.”
It states the timeframe, observation, interpretation, and invalidation. “Looks bullish” states none of them.
### 4. Check participation: volume, breadth, and spot demand [#4-check-participation-volume-breadth-and-spot-demand]
Participation asks whether more activity is supporting the price move.
Compare:
* Spot volume with its own trailing baseline, not with an arbitrary universal threshold
* Number of advancing assets in the relevant sector
* Performance across several credible venues
* Spot movement versus perpetual-futures movement
* Protocol usage or fee activity when the token thesis depends on usage
Volume is activity, not direction, and provider methodologies can differ. A single exchange can also be unrepresentative. Record which venues are included and whether the number is reported or adjusted.
Use a **confirmation table**:
| Observation | Confirms the move? | Alternative explanation | How to verify |
| ----------------------- | ------------------ | ---------------------------- | ------------------------------------- |
| Price breaks a range | Not by itself | Thin weekend book | Compare depth and venue breadth |
| Spot volume expands | Partly | One-off listing or transfer | Inspect venues and sustained activity |
| Sector breadth improves | Yes, if persistent | Index composition changed | Use a stable comparison set |
| Protocol fees rise | Maybe | Temporary incentive campaign | Compare fees, users, and incentives |
The alternative-explanation column is the defense against confirmation bias.
### 5. Separate leverage from conviction [#5-separate-leverage-from-conviction]
Open interest is the number of derivatives contracts that remain open, not a count of bullish positions. The CFTC defines it as contracts entered into and not yet offset, delivered, or exercised; aggregate long open interest equals aggregate short open interest ([CFTC explanatory notes](https://www.cftc.gov/MarketReports/CommitmentsofTraders/ExplanatoryNotes/index.htm)). CME likewise distinguishes volume, which counts traded contracts, from open interest, which counts contracts still open ([CME Group](https://www.cmegroup.com/education/courses/introduction-to-futures/open-interest)).
Therefore:
* Rising price plus rising open interest means more positions are open; it does **not** identify who is right.
* Funding can indicate which side pays to maintain perpetual exposure, but it varies by venue.
* Liquidations show forced position closures, not the underlying reason people opened positions.
* A basis or funding extreme can persist, so “crowded” is a risk description, not a timing signal.
Write leverage observations in pairs:
> “Price and aggregate open interest rose together, while funding became more positive. This is consistent with added long-side demand but also greater liquidation sensitivity. Spot-volume breadth must confirm the move.”
The first sentence interprets; the second limits the inference.
### 6. Test liquidity and executable reality [#6-test-liquidity-and-executable-reality]
Market capitalization and 24-hour volume do not tell you what a realistic order would receive. Inspect:
* Bid-ask spread
* Order-book depth at fixed percentages from mid-price
* Expected price impact at several order sizes
* DEX pool liquidity in the active price range
* Distribution across venues and pairs
* Whether a team or a small wallet set controls removable liquidity
Uniswap defines price impact as the price change caused directly by a trade and notes that lower pool liquidity generally produces larger impact ([Uniswap Labs](https://support.uniswap.org/hc/en-us/articles/8671539602317-What-is-Price-Impact)). Coinbase similarly warns that a market order may fill away from the last trade when available orders are insufficient ([Coinbase](https://help.coinbase.com/en/coinbase/trading-and-funding/advanced-trade/order-management)).
Run the complete process in [Crypto Liquidity Analysis](/blog/crypto-liquidity-analysis). The essential principle is simple: **measure liquidity at the size relevant to your question**. “Liquid” is not a permanent property of a ticker.
### 7. Add catalysts, fundamentals, and on-chain evidence [#7-add-catalysts-fundamentals-and-on-chain-evidence]
Only after describing the market should you ask why it could continue.
For a protocol token, check:
* Product usage, fees, and revenue definitions
* Token supply, emissions, and [upcoming unlocks](/blog/token-unlocks-vesting-guide)
* Governance or protocol changes
* Competitive changes
* Security incidents, audit updates, and privileged contract controls
* [Whale and exchange flows](/blog/whale-wallets-exchange-flows), with label confidence disclosed
Do not promote a calendar item to a causal story without evidence. A scheduled upgrade can coincide with price strength without causing it. Write catalysts as scenarios:
> “The upgrade is a potential catalyst. Confirmation would be sustained usage or fee changes after launch; counterevidence would be unchanged usage despite the release.”
## A decision matrix for combining the layers [#a-decision-matrix-for-combining-the-layers]
Do not average incompatible data into a magical score. Use a matrix that preserves disagreement:
| Layer | Constructive | Neutral/unclear | Adverse | Confidence |
| ------------- | ------------------------ | --------------- | ----------------- | --------------- |
| Regime | Broad expansion | Transition | Contraction | Low/medium/high |
| Structure | Higher-timeframe advance | Range | Sustained decline | Low/medium/high |
| Participation | Broad spot confirmation | Mixed | Narrow or fading | Low/medium/high |
| Leverage | Balanced | Unclear | One-sided/fragile | Low/medium/high |
| Liquidity | Deep across venues | Uneven | Thin/concentrated | Low/medium/high |
| Fundamentals | Improving with evidence | Stable | Deteriorating | Low/medium/high |
Then write:
1. **Base interpretation:** the explanation best supported now
2. **Strongest contradiction:** the layer that disagrees
3. **Invalidation:** observable evidence that changes the interpretation
4. **Next check date or event:** when the analysis should be refreshed
If five weak data sources agree, that is not necessarily stronger than one primary source that disproves them.
## Worked hypothetical example: analyzing “Northstar” [#worked-hypothetical-example-analyzing-northstar]
**This example is fictional. It is not a real token, market observation, recommendation, or backtest.**
Question: Has Northstar’s six-week advance become durable enough for a full protocol review?
| Layer | Hypothetical observation | Interpretation | Counterevidence needed |
| ------------- | ------------------------------------------------------------- | ---------------------------------------------------- | --------------------------------------------------------- |
| Regime | BTC trend positive; sector breadth mixed | Backdrop supportive but narrow | Breadth continues falling |
| Structure | Weekly range reclaimed; TOKEN/BTC still below prior high | USD structure improved; relative strength incomplete | Weekly close back inside range |
| Participation | Spot volume 1.6× its 90-day median across three venues | Better-than-usual participation | Expansion disappears after two sessions |
| Leverage | Open interest +35%; funding positive | Added exposure with higher fragility | Spot demand fades as OI rises |
| Liquidity | Tight top-of-book spread; material impact at larger test size | Small orders feasible; larger size constrained | Depth improves across venues |
| Fundamentals | Fees rising, but a 9% circulating-supply unlock is scheduled | Usage evidence conflicts with supply risk | Unlock recipients retain or delegate rather than transfer |
Base interpretation: **constructive but unconfirmed**. Structure and participation improved, while relative strength, leverage, liquidity, and the unlock limit confidence.
Invalidation: a weekly close back inside the prior range alongside fading spot volume.
Next research action: run the [token unlock analysis](/blog/token-unlocks-vesting-guide), verify fee methodology, and recheck market depth at three standardized sizes.
Notice what the conclusion does not say. It does not predict a price, turn six mixed inputs into “82% bullish,” or infer that unlocked holders will sell.
## A reproducible crypto market analysis checklist [#a-reproducible-crypto-market-analysis-checklist]
Copy this into your research note:
```text
Question:
Asset / reference pairs:
Research window:
Decision window:
Data timestamp and timezone:
REGIME
- Policy-rate direction and dated events:
- BTC structure:
- BTC dominance / market breadth:
- Sentiment context:
STRUCTURE
- Higher-timeframe label:
- Major accepted/rejected areas:
- Relative-strength result:
- Structural invalidation:
PARTICIPATION
- Spot volume vs trailing baseline:
- Venues included:
- Sector breadth:
- Usage/fee evidence:
LEVERAGE
- Open-interest change:
- Funding/basis across venues:
- Liquidation sensitivity:
- What this data cannot prove:
LIQUIDITY
- Spread:
- Depth at standardized distances:
- Price impact at relevant sizes:
- Venue/pool concentration:
CATALYSTS AND FUNDAMENTALS
- Dated catalysts:
- Supply changes/unlocks:
- Security/governance changes:
- On-chain evidence and label confidence:
Base interpretation:
Strongest contradiction:
Invalidation:
Next review trigger:
Sources and retrieval times:
```
For a full start-to-finish process, use the [Crypto Research Workflow Template](/blog/crypto-research-workflow-template). To decide which tasks should be automated, see [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research).
## Common crypto market analysis mistakes [#common-crypto-market-analysis-mistakes]
### Treating correlation as a cause [#treating-correlation-as-a-cause]
Price rose after news; therefore news caused price. That is a hypothesis, not a verified causal chain. Check timing, market-wide movement, flow evidence, and whether the story appeared after the move.
### Mixing spot, perpetual, and futures definitions [#mixing-spot-perpetual-and-futures-definitions]
They measure different markets. Label venue, instrument, collateral, and interval. Do not merge them without a documented methodology.
### Comparing unlike protocols [#comparing-unlike-protocols]
TVL, users, and revenue do not have identical meanings across exchanges, lenders, bridges, and base layers. Compare business mechanisms before ratios.
### Using today’s price in a fully diluted valuation as if it were a forecast [#using-todays-price-in-a-fully-diluted-valuation-as-if-it-were-a-forecast]
FDV is a scenario calculation, not the amount of future capital guaranteed to enter the asset. Read [Market Cap vs FDV](/blog/market-cap-vs-fdv-crypto) before using it.
### Hiding uncertainty in a score [#hiding-uncertainty-in-a-score]
A score can improve consistency, but it can also hide weak inputs. Preserve the source and contradiction behind every label.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Market data can be revised, delayed, fragmented, or defined differently across providers.
* On-chain transfers reveal addresses and transactions, not private intent. Entity labels can be incomplete.
* Technical structure is sensitive to timeframe and chosen reference pair.
* Macro relationships change; historical correlation is not a law.
* Public fundamentals may omit off-chain activity or count incentivized activity as organic demand.
* An internally consistent analysis can still be wrong because markets respond to new information.
These limits are reasons to retain timestamps, sources, and invalidations—not reasons to abandon analysis.
## How BlockMind can support the workflow [#how-blockmind-can-support-the-workflow]
A BlockMind agent can collect current market data, research public sources, compare on-chain and market evidence, and save the resulting framework in your Notebook. Its [research capabilities](/docs/capabilities/research) cover asset, narrative, equities, and news research; [on-chain intelligence](/docs/capabilities/on-chain) adds holder, wallet, flow, TVL, fee, and DEX context. You can also ask it to revisit live protocol pages with its [browser capability](/docs/capabilities/browser).
The useful division of labor is mechanical collection and monitoring by your agent, followed by human review of assumptions, source quality, invalidation, and risk. Always [verify an AI-generated crypto analysis](/blog/how-to-verify-ai-crypto-analysis) before relying on it.
## The Bottom Line [#the-bottom-line]
Crypto market analysis becomes more reliable when it becomes reproducible. State the question and horizon, work from regime to structure, verify participation, separate leverage from conviction, test executable liquidity, then connect catalysts to measurable fundamentals. Preserve the strongest contradiction and the evidence that would make you change your mind.
The goal is not certainty. It is a research record that another careful reader can inspect, challenge, and update.
## Sources [#sources]
* [Federal Reserve — Economy at a Glance: Policy Rate](https://www.federalreserve.gov/economy-at-a-glance-policy-rate.htm)
* [CFTC — Commitments of Traders Explanatory Notes](https://www.cftc.gov/MarketReports/CommitmentsofTraders/ExplanatoryNotes/index.htm)
* [CME Group — Open Interest](https://www.cmegroup.com/education/courses/introduction-to-futures/open-interest)
* [Uniswap Labs — What Is Price Impact?](https://support.uniswap.org/hc/en-us/articles/8671539602317-What-is-Price-Impact)
* [Coinbase — Order Management and Slippage](https://help.coinbase.com/en/coinbase/trading-and-funding/advanced-trade/order-management)
* [CFTC — Understand the Risks of Virtual Currency Trading](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/understand_risks_of_virtual_currency.html)
---
# Crypto Morning Brief: A Daily Market Briefing Written About Your Portfolio
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: informational
Primary keyword: crypto morning brief
Keywords: crypto morning brief, daily crypto briefing, crypto portfolio monitoring
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A useful crypto morning brief is not a list of prices or yesterday's headlines. It connects overnight events to your portfolio, tells you what changed, and separates what needs attention from what can wait. The value comes from relevance and consistency, not length.
***
Crypto never closes. By the time you wake up, prices may have moved, funding may have shifted, a protocol may have published an update, and a token you follow may have crossed a level you cared about.
Most investors respond by opening five apps before breakfast. A crypto morning brief is a better interface: one prepared summary of what happened, why it matters to you, and what deserves a closer look.
## What is a crypto morning brief? [#what-is-a-crypto-morning-brief]
A crypto morning brief is a short daily market briefing prepared around an investor's holdings, watchlists, and open research.
It should answer four questions:
1. What changed while I was away?
2. Why did it change?
3. Which changes matter to my portfolio?
4. What should I watch today?
The first question is easy. Every price app can show a percentage. The other three require context.
## What belongs in a useful morning brief? [#what-belongs-in-a-useful-morning-brief]
### One clear headline [#one-clear-headline]
Start with the market fact that changes the day's context. That could be a broad risk-off move, a sharp shift in Bitcoin dominance, an exploit, a macro release, or a portfolio-specific event.
If nothing important changed, the brief should say so. Manufacturing urgency every morning trains the reader to ignore it.
### Portfolio movers with reasons [#portfolio-movers-with-reasons]
A list of green and red percentages is not analysis. Each material mover should include a short explanation and an honest confidence level.
Good:
> SOL is up 8% after a combination of market-wide risk appetite and renewed ecosystem activity. Your position is now above the level you marked as confirmation, but volume has not expanded enough to settle the question.
Weak:
> SOL +8%. Bullish momentum.
The difference is evidence, portfolio context, and uncertainty.
### Market regime [#market-regime]
Individual tokens behave differently in different regimes. A useful brief checks broader context such as:
* The [Fear & Greed Index](/blog/fear-greed-index-guide)
* [Bitcoin dominance](/blog/bitcoin-dominance-explained)
* Altcoin breadth and sector rotation
* Funding, open interest, and leverage
* Stablecoin or liquidity stress
These indicators are context, not trade signals. Their job is to explain the weather around your positions.
### Today’s catalysts and risks [#todays-catalysts-and-risks]
The brief should look forward as well as backward. Relevant items might include:
* Token unlocks and governance votes
* Earnings, economic releases, or regulatory events
* Protocol upgrades and product launches
* Levels that validate or invalidate an open thesis
* Scheduled announcements from teams you follow
The list should be short enough that each item earns attention.
### A path to go deeper [#a-path-to-go-deeper]
Every important claim should lead somewhere. A reader should be able to open the supporting source, continue the research, or ask a follow-up without rebuilding the context from scratch.
## A morning brief is not a newsletter [#a-morning-brief-is-not-a-newsletter]
Newsletters can be excellent, but they are written for an audience. A personal brief is written for one portfolio.
| Newsletter | Personal Morning Brief |
| ----------------------------------- | ------------------------------------------ |
| Covers the editor's chosen stories | Prioritizes your holdings and watchlists |
| Same version for every reader | Changes with your exposure and open theses |
| Usually explains the market broadly | Explains what the market means for you |
| Arrives on the publisher's schedule | Arrives on your local schedule |
| Starts fresh each issue | Can remember prior research and decisions |
The two can complement each other. A strong newsletter expands your perspective. A strong personal brief reduces the chance that relevant information slips through.
## A morning brief is not an alert feed [#a-morning-brief-is-not-an-alert-feed]
Alerts answer "did a condition happen?" A brief answers "what does the current picture mean?"
An alert is useful when BTC crosses a level, a holding moves sharply, or a protocol reports an exploit. The morning brief is useful when several smaller changes combine into a new context.
Flooding a brief with every overnight notification defeats the point. The brief should compress, compare, and rank.
## Why portfolio context changes the analysis [#why-portfolio-context-changes-the-analysis]
Consider a rise in Bitcoin dominance.
For someone holding mostly Bitcoin, that may reinforce the current positioning. For someone concentrated in small-cap altcoins, the same move can signal a harder environment even if the total portfolio is still green. For a third investor holding stablecoins and waiting for an entry, it may simply be context.
The indicator has not changed. The meaning has.
This is why a generic market recap cannot become personal by inserting the reader's name. The system needs the actual portfolio, the watchlist, and the reasons behind the positions.
## What a Morning Brief should never do [#what-a-morning-brief-should-never-do]
A credible brief should not:
* Turn every observation into a buy or sell instruction
* Hide uncertainty behind confident language
* Claim a cause when only a correlation is known
* Treat a once-daily indicator as real-time data
* Include private keys, seed phrases, or unnecessary account data
* Repeat the same generic market summary regardless of the portfolio
* Bury the source of a material claim
The point is better judgment, not more activity.
## How to read a crypto brief in five minutes [#how-to-read-a-crypto-brief-in-five-minutes]
Use a simple sequence:
1. **Read the headline.** Did the market context materially change?
2. **Scan your movers.** Is the move explained, and is the explanation sourced?
3. **Check invalidations.** Did anything challenge the reason you hold or follow an asset?
4. **Note today's catalysts.** Which event could require attention later?
5. **Choose one follow-up.** Go deeper only where the expected value is highest.
If the brief regularly sends you into ten new tabs, it is not doing enough compression.
## How BlockMind's Morning Brief works [#how-blockminds-morning-brief-works]
By default, a BlockMind Pro agent prepares a Morning Brief at about 8:00 in the user's timezone. The brief is published to the agent dashboard first and sent to the user's email inbox.
The brief is grounded in connected portfolios, watchlists, tracked research, and the agent's Notebook. It can include:
* The most important overnight development
* Material moves among holdings and tracked assets
* Plain-language reasons behind those moves
* Portfolio-specific implications
* Events, risks, and levels worth watching that day
* Follow-up prompts that continue the research in web chat
You can change the emphasis in plain language. Ask for more macro context, less coverage of a sector, shorter briefs, or a recurring section tied to a saved verdict. The agent keeps that preference for future briefs.
Morning Brief delivery is dashboard plus email. Other channels may support chat or alerts when available, but the daily brief should not be described as a Telegram delivery.
## Frequently Asked Questions [#frequently-asked-questions]
### What time should a crypto morning brief arrive? [#what-time-should-a-crypto-morning-brief-arrive]
Just before the reader normally reviews the market. BlockMind currently configures the brief for about 8:00 in the user's local timezone.
### How long should a daily crypto brief be? [#how-long-should-a-daily-crypto-brief-be]
Long enough to explain the few changes that matter, short enough to scan in several minutes. Relevance is a better constraint than a fixed word count.
### Is the Fear & Greed Index enough for a morning brief? [#is-the-fear--greed-index-enough-for-a-morning-brief]
No. It is one sentiment input. Use it with price structure, Bitcoin dominance, leverage, news, on-chain evidence, and portfolio context. Never treat one indicator as a decision.
### Can a morning brief tell me what to buy? [#can-a-morning-brief-tell-me-what-to-buy]
It can surface candidates and explain evidence, but the final decision should stay with you. A brief should improve the starting point for judgment, not replace judgment.
### Does BlockMind send the Morning Brief through Telegram? [#does-blockmind-send-the-morning-brief-through-telegram]
No. The Morning Brief is published to the dashboard and sent by email. Telegram, when available and connected, is a separate companion for chat and alerts.
### Do I need the BlockMind agent for free portfolio tracking? [#do-i-need-the-blockmind-agent-for-free-portfolio-tracking]
No. The classic platform remains free for portfolio tracking, watchlists, charts, and market indicators within its limits. The personal agent and Morning Brief come with Pro.
## The Bottom Line [#the-bottom-line]
A crypto morning brief should save attention, not create another feed. It earns its place by knowing what you hold, remembering what you care about, and explaining only the changes that could alter your judgment.
The best brief does not make you trade more. It helps you begin the day with fewer blind spots and a clearer idea of what can safely wait.
BlockMind's agent remains read-only: it cannot trade, withdraw, transfer, or sign transactions. The brief provides research context, not individualized financial advice.
## Learn More [#learn-more]
* [How BlockMind Morning Briefs work](/docs/automations/morning-brief)
* [What Is an AI Crypto Agent?](/blog/what-is-an-ai-crypto-agent)
* [AI Portfolio Monitoring](/blog/ai-portfolio-monitoring-crypto)
* [Bitcoin Dominance Explained](/blog/bitcoin-dominance-explained)
* [How to Read the Crypto Fear & Greed Index](/blog/fear-greed-index-guide)
## Sources [#sources]
1. [Federal Reserve: FOMC calendars and information](https://www.federalreserve.gov/monetarypolicy/fomccalendars.htm), accessed July 2026.
2. [U.S. Bureau of Labor Statistics: Release calendar](https://www.bls.gov/schedule/), accessed July 2026.
3. [Alternative.me: Crypto Fear & Greed Index methodology](https://alternative.me/crypto/fear-and-greed-index/), accessed July 2026.
---
# Crypto Portfolio Concentration Risk: Four Ways to Measure Hidden Exposure
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: informational
Primary keyword: crypto portfolio concentration risk
Keywords: crypto portfolio concentration risk, crypto diversification, portfolio concentration calculator, crypto portfolio allocation risk, hidden crypto exposure
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Count crypto concentration four ways: largest position weights, the portfolio-wide Herfindahl-Hirschman Index, look-through exposure to shared ecosystems and infrastructure, and position size relative to executable liquidity. A portfolio can hold many tokens and still depend on one asset, chain, stablecoin, bridge, custodian, or market regime.
***
Concentration risk is the possibility of amplified loss because a large part of a portfolio depends on the same investment, asset class, or market segment. FINRA notes that concentration can be intentional, emerge after one holding outperforms, arise through correlated assets, or hide inside overlapping investments.
Concentration is one layer of the broader [AI portfolio monitoring framework](/blog/ai-portfolio-monitoring-crypto), where holdings, theses, and material changes remain connected over time.
Crypto makes the “look under the hood” problem unusually important. A wallet may show ten tickers, but four can be claims on ETH, three can rely on one chain and bridge, and the rest can share the same stablecoin collateral or centralized custodian.
This guide measures exposure. It does not prescribe an ideal allocation or tell you to rebalance.
## Measure 1: top-position weights [#measure-1-top-position-weights]
For each position:
`position weight = position value ÷ total portfolio value × 100`.
Then calculate:
* **Top 1:** largest individual exposure;
* **Top 3:** combined weight of the three largest exposures;
* **Top 5:** combined weight of the five largest exposures.
### Hypothetical portfolio [#hypothetical-portfolio]
| Position | Marked value | Portfolio weight |
| ------------------- | -----------: | ---------------: |
| BTC | $40,000 | 40% |
| ETH | $20,000 | 20% |
| Staked ETH receipt | $12,000 | 12% |
| Layer-2 token | $10,000 | 10% |
| USDC | $8,000 | 8% |
| DeFi protocol token | $6,000 | 6% |
| Memecoin | $4,000 | 4% |
| **Total** | **$100,000** | **100%** |
Top 1 is 40%. Top 3 is `40 + 20 + 12 = 72%`. These numbers are easy to communicate but depend on how positions are classified. ETH and the staked ETH receipt are separate rows even though they share underlying exposure.
No universal cutoff makes a portfolio healthy or unhealthy. A threshold should come from the owner's objectives, risk capacity, time horizon, tax constraints, and liquidity needs. The metric's job is to make the choice visible.
## Measure 2: HHI across positions [#measure-2-hhi-across-positions]
The Herfindahl-Hirschman Index adds the squared decimal weights of every position:
`HHI = sum of each position weight squared`.
Using decimal weights for the hypothetical portfolio:
`0.40² + 0.20² + 0.12² + 0.10² + 0.08² + 0.06² + 0.04² = 0.236`.
Convert that into an “effective number of equally weighted positions”:
`effective positions = 1 ÷ HHI`.
Here, `1 ÷ 0.236 = 4.24`. The wallet has seven ticker rows, but its weight concentration resembles roughly 4.24 equally weighted positions.
HHI is a descriptive concentration measure, not a loss forecast. It assumes rows are independent categories, which they are not. That is why look-through mapping comes next.
## Measure 3: look-through dependency exposure [#measure-3-look-through-dependency-exposure]
Create a matrix linking every position to the underlying systems it depends on.
| Position | Underlying asset | Chain | Protocol/admin | Bridge/oracle | Custodian/venue |
| ------------------ | ------------------------ | ------------- | ----------------- | ---------------------------- | ---------------------- |
| ETH | ETH | Ethereum | Ethereum protocol | Native feeds/venues | Wallet/exchange |
| Staked ETH receipt | ETH + staking derivative | Ethereum | Staking protocol | Protocol oracle | Smart contract |
| Layer-2 token | L2 ecosystem | L2 + Ethereum | L2 governance | Canonical bridge | Wallet/exchange |
| USDC | USD stablecoin | Multiple | Issuer controls | Bridges on non-native chains | Issuer/custodian stack |
Now aggregate weights for each dependency. Do not force weights to sum to 100%; one position can depend on several systems.
### Worked look-through example [#worked-look-through-example]
Suppose you classify the staked ETH receipt's entire $12,000 as underlying ETH exposure and the layer-2 token's $10,000 as Ethereum-ecosystem exposure, without claiming it is equivalent to ETH price exposure.
* Direct ETH + staked ETH underlying = `$20,000 + $12,000 = $32,000`, or 32%.
* Broader Ethereum ecosystem including the L2 token = `$32,000 + $10,000 = $42,000`, or 42%.
This does **not** mean the three positions will move identically. It means 42% depends materially on some part of the Ethereum ecosystem. The staked receipt also adds protocol and smart-contract risk that direct ETH may not share.
Use separate columns for:
* directional price exposure;
* chain availability;
* smart-contract dependency;
* bridge dependency;
* oracle dependency;
* stablecoin issuer;
* exchange or custodian;
* governance/admin key;
* narrative or sector.
This prevents one label from collapsing different risks.
## Measure 4: liquidity-adjusted concentration [#measure-4-liquidity-adjusted-concentration]
A 4% position can be more operationally concentrated than a 20% position if it is difficult to sell.
Calculate:
`liquidity-adjusted exposure = position value ÷ executable depth within chosen price impact`.
For the hypothetical $4,000 memecoin, suppose only $1,000 can be sold within a 10% price-impact band. Liquidity-adjusted exposure is `4,000 ÷ 1,000 = 4`.
Suppose the $40,000 BTC position has $2 million of depth on the selected venue within the same band. Its ratio is `40,000 ÷ 2,000,000 = 0.02`.
The memecoin is only 4% of marked portfolio value but is four times the measured exit depth. BTC is 40% by weight but small relative to that venue's quoted depth. These measure different risks: BTC dominates directional exposure; the memecoin may dominate exit friction.
Depth changes across venues and time. Record source, timestamp, pairs, fees, and price band. Never treat a public aggregate as guaranteed execution.
## Recalculate HHI after look-through grouping [#recalculate-hhi-after-look-through-grouping]
The original HHI treats ETH, the staking receipt, and L2 token as separate. To stress a shared ecosystem, group them:
| Group | Weight |
| ------------------ | -----: |
| BTC ecosystem | 40% |
| Ethereum ecosystem | 42% |
| USDC | 8% |
| Other DeFi | 6% |
| Memecoin | 4% |
Grouped HHI is:
`0.40² + 0.42² + 0.08² + 0.06² + 0.04² = 0.348`.
Effective grouped positions are `1 ÷ 0.348 = 2.87`.
Classification changed the story from 4.24 effective ticker positions to 2.87 effective ecosystem groups. Neither number is “the truth.” Together they expose how much the conclusion depends on the risk lens.
## Seven types of crypto concentration [#seven-types-of-crypto-concentration]
### 1. Asset concentration [#1-asset-concentration]
One coin or token dominates marked value.
### 2. Ecosystem concentration [#2-ecosystem-concentration]
Several assets rely on the same chain, rollup, staking layer, or application ecosystem.
### 3. Stablecoin concentration [#3-stablecoin-concentration]
Multiple pairs, lending positions, and liquidity pools may depend on one stablecoin issuer and redemption mechanism.
### 4. Protocol concentration [#4-protocol-concentration]
Collateral, yield, LP positions, and receipt tokens can all depend on one protocol's contracts and governance.
### 5. Infrastructure concentration [#5-infrastructure-concentration]
Different tokens may share an oracle, bridge, sequencer, wallet, or RPC dependency.
### 6. Custody concentration [#6-custody-concentration]
Assets across several tickers can sit on one exchange or under one signing setup. If access fails, the whole group is affected.
### 7. Liquidity concentration [#7-liquidity-concentration]
Many positions may depend on the same base asset, pool, market maker, or venue for exit.
Count each explicitly in a [crypto portfolio health check](/blog/crypto-portfolio-health-check).
For token-specific exposure, pair the portfolio view with [holder-distribution analysis](/blog/crypto-holder-distribution) so a diversified wallet list does not hide one concentrated issuer or insider base.
## Correlation is useful—and unstable [#correlation-is-usefuland-unstable]
Historical correlation can show which assets moved together, but it is not a permanent property. Relationships can strengthen during market stress. New tokens have short histories, and sparse pricing can make correlations misleading.
Use correlation as supporting evidence, not a complete concentration map. Two assets can share a bridge or admin key even if their prices were uncorrelated. Two tokens can become highly correlated during a selloff despite different narratives.
A scenario approach is often clearer:
* What breaks if BTC falls sharply?
* What breaks if Ethereum finality or a major L2 is disrupted?
* What breaks if one stablecoin loses its peg?
* What breaks if a bridge, oracle, or custodian pauses?
* What breaks if DEX liquidity contracts by half?
* What breaks if a governance key is compromised?
Map impacted positions and value for each scenario without predicting its probability.
## Concentration created by performance drift [#concentration-created-by-performance-drift]
A position can become large because it appreciated faster than the rest of the portfolio. Investor.gov notes that this can move an allocation away from the investor's goals and change risk. FINRA recommends periodic review and looking through funds for overlap; the same logic applies to token wrappers and protocols.
Track:
`weight drift = current weight - intended/reference weight`.
**Hypothetical:** An exposure began at 15% and grew to 27%. Drift is `27 - 15 = 12 percentage points`.
This does not say what action to take. Rebalancing can create taxes, fees, slippage, and loss of desired exposure. It says the current risk is no longer the one originally documented.
Use [crypto alerts beyond price](/blog/crypto-alerts-beyond-price) to turn material weight drift, dependency changes, and liquidity deterioration into research triggers rather than automatic trades.
## A reproducible concentration worksheet [#a-reproducible-concentration-worksheet]
* [ ] Choose one portfolio timestamp.
* [ ] Reconcile all wallets, exchanges, protocols, and debt.
* [ ] Calculate each marked-value weight.
* [ ] Calculate top 1, top 3, and top 5.
* [ ] Calculate raw-position HHI and effective positions.
* [ ] Map underlying asset and ecosystem dependencies.
* [ ] Map stablecoin, bridge, oracle, protocol, admin, and custody dependencies.
* [ ] Recalculate grouped HHI under at least two classification schemes.
* [ ] Test exit depth for every material or illiquid position.
* [ ] Run shared-failure scenarios.
* [ ] Record intentional versus accidental concentration.
* [ ] Save data sources, timestamp, unknowns, and review triggers.
## Limitations and counterevidence [#limitations-and-counterevidence]
Diversification does not guarantee profit or prevent loss. Holding more tokens can add complexity, fees, attack surface, and low-quality exposure without reducing the dominant risk. Concentration can be deliberate and consistent with an owner's goals; a metric cannot decide suitability.
Marked values, holder classifications, liquidity, and correlations can be wrong or stale. Wrapped and derivative assets require protocol-specific treatment. HHI does not understand dependencies unless you group them, and grouping introduces analyst judgment.
Tax, legal, household-balance-sheet, and cash-flow context may change the interpretation. Consult qualified professionals where appropriate.
## How BlockMind can help [#how-blockmind-can-help]
A connected BlockMind portfolio gives your agent balance and position context for allocation, exposure, performance, and change analysis. Ask it to calculate raw weights, map look-through dependencies, document classification assumptions, and save the analysis in the [Notebook](/docs/memory/notebook).
Verify calculations against the source accounts and protocol data. The agent can be wrong, does not know every hidden dependency, and cannot decide an appropriate allocation for you. BlockMind cannot trade or move funds.
## The Bottom Line [#the-bottom-line]
Ticker count is not diversification. Measure top weights and HHI, then look through the wrappers and group shared dependencies. Finally, compare marked position size with actual exit depth.
The most valuable concentration finding is often not “you own too much X.” It is “these positions you thought were different depend on the same thing.”
This article is for research and education, not financial, tax, or legal advice.
## Sources [#sources]
1. [FINRA: Concentration risk and correlated holdings](https://www.finra.org/investors/insights/concentration-risk)
2. [FINRA: Asset allocation and diversification](https://www.finra.org/investors/investing/investing-basics/asset-allocation-diversification)
3. [FINRA: Investment and liquidity risk](https://www.finra.org/investors/investing/investing-basics/risk)
4. [Investor.gov: Asset allocation, diversification, and rebalancing](https://www.investor.gov/introduction-investing/getting-started/asset-allocation)
5. [BlockMind: Portfolio analysis](/docs/capabilities/portfolio-analysis)
6. [BlockMind: Portfolio data model](/docs/data/portfolios)
---
# Crypto Portfolio Health Check: A Six-Layer Review You Can Audit
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: problem-solving
Primary keyword: crypto portfolio health check
Keywords: crypto portfolio health check, crypto portfolio analysis, crypto risk assessment, review crypto portfolio, portfolio health checklist
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A crypto portfolio health check should answer six questions: Are the holdings complete? Where is the real exposure? What can fail operationally? How liquid are the positions? Which theses have changed? And will the monitoring surface the changes that matter? It is a diagnostic process, not a score that tells you what to buy or sell.
***
A portfolio can look healthy on a pie chart and still hide missing accounts, duplicated exposure, shallow liquidity, stale cost basis, dangerous permissions, or an investment thesis that no longer matches reality.
For the cluster's monitoring model, read [AI Portfolio Monitoring for Crypto](/blog/ai-portfolio-monitoring-crypto). This health check turns that ongoing monitoring problem into a dated, auditable review.
The goal of a health check is not to force every portfolio toward the same allocation. Investor.gov emphasizes that asset allocation depends on a person's time horizon and risk tolerance. The useful output is a dated map of what you own, what can hurt it, what evidence supports each thesis, and what needs to be watched.
## The HEALTH framework [#the-health-framework]
Use six layers:
1. **H — Holdings integrity**
2. **E — Exposure and concentration**
3. **A — Access, custody, and permissions**
4. **L — Liquidity and liabilities**
5. **T — Thesis and time horizon**
6. **H — Habits, alerts, and review cadence**
Run the check against a single timestamp. Crypto balances, prices, positions, and protocol states change continuously; mixing screenshots from different days can create false reconciliation breaks.
## H — Holdings integrity [#h--holdings-integrity]
Start by proving the portfolio view is complete and internally consistent.
### Build a source inventory [#build-a-source-inventory]
List every location where exposure exists:
* centralized exchanges;
* public wallets on each chain;
* hardware and multisig wallets;
* staked and liquid-staked positions;
* lending collateral and debt;
* liquidity-pool positions;
* bridges and wrapped assets;
* vesting or locked tokens;
* manual or off-platform positions;
* derivatives, if applicable;
* fiat and stablecoin balances relevant to the portfolio.
Then record each source's last sync time and method: public address, read-only exchange connection, CSV, or manual entry.
### Calculate a completeness ratio [#calculate-a-completeness-ratio]
Use reconciled value, not number of accounts:
`portfolio completeness = reconciled source value ÷ independently estimated total value × 100`.
**Hypothetical:** The portfolio dashboard reconciles $92,000. A source inventory identifies another $5,000 in a staking contract and $3,000 on an unsupported exchange. Estimated total value is $100,000, so completeness is `92,000 ÷ 100,000 × 100 = 92%`.
That is not a quality score for the assets. It means 8% of estimated value is absent from the analytical view. Any concentration or performance calculation using the incomplete view is provisional.
### Reconcile common errors [#reconcile-common-errors]
* duplicate wallets imported through two paths;
* spam or dust tokens assigned misleading prices;
* bridged assets counted on both source and destination;
* debt omitted while collateral is included;
* staking receipt token counted alongside underlying assets;
* missing internal exchange transfers;
* stale or missing cost basis;
* token symbols mapped to the wrong contract;
* delisted or illiquid tokens shown at an unexecutable price.
BlockMind's [portfolio documentation](/docs/data/portfolios) explains the differences among wallet-linked, exchange-connected, CSV, and manual portfolios, plus spam filtering. Any tracker can inherit limitations from the source API. Reconciliation remains necessary.
## E — Exposure and concentration [#e--exposure-and-concentration]
Count economic exposures, not ticker rows.
A portfolio with ETH, a liquid-staking token, an ETH lending position, and an ETH-correlated layer-2 token may look diversified by name while remaining dependent on the same underlying asset and ecosystem. FINRA defines concentration risk as amplified loss from a large share in one investment, asset class, or market segment and notes that correlated holdings can create hidden concentration.
Review at least these lenses:
* top one, three, and five assets;
* chain and ecosystem;
* protocol and smart-contract dependency;
* stablecoin issuer and reserve model;
* centralized exchange or custodian;
* bridge and oracle;
* narrative or sector;
* collateral asset;
* liquidity venue;
* common governance or admin key;
* directional beta to BTC or ETH;
Use [Crypto Portfolio Concentration Risk](/blog/crypto-portfolio-concentration-risk) for formulas including weights, HHI, and look-through exposure.
Do not declare a universal “healthy” percentage. A deliberate high-conviction portfolio, a short-horizon trading book, and long-term household savings have different constraints. Record whether each concentration is intentional, understood, and consistent with the user's actual ability to bear loss.
## A — Access, custody, and permissions [#a--access-custody-and-permissions]
Operational health is part of portfolio health.
Inventory:
* who or what can sign each wallet;
* backup and recovery arrangements;
* hardware-wallet and multisig use;
* exchange MFA and withdrawal controls;
* API keys and their permissions;
* token approvals and delegated authority;
* protocol admin, oracle, bridge, and custody dependencies;
* single points of failure;
* inheritance or emergency-access plans where relevant.
For portfolio tracking, use public addresses or provider-scoped read-only credentials where available. The [read-only API key guide](/blog/read-only-api-keys-crypto-security) explains that read-only limits financial authority but does not remove privacy, credential, or third-party risk.
BlockMind reads balances and positions and cannot initiate trades, withdrawals, transfers, or wallet signatures. The Firi key-scope exception and other supported-provider details live in [Connect your portfolio](/docs/getting-started/connect-your-portfolio).
## L — Liquidity and liabilities [#l--liquidity-and-liabilities]
Displayed value is not necessarily realizable value.
For each material position, record:
* venue and pair;
* quoted depth at realistic sizes;
* spread, slippage, gas, and withdrawal constraints;
* lock, vesting, or unbonding period;
* collateral and liquidation threshold;
* debt denomination and rate type;
* bridge or redemption route;
* market hours or operational dependencies;
* whether a price source reflects an active market.
### Run a liquidity stress worksheet [#run-a-liquidity-stress-worksheet]
For each position calculate:
`liquidity stress ratio = position value ÷ executable depth within chosen price impact`.
**Hypothetical:** A token position is displayed at $30,000. A depth tool estimates only $10,000 can be sold within 10% price impact. The ratio is `30,000 ÷ 10,000 = 3`. The marked position is three times the measured depth within that band.
This is not a forecast or universal threshold. Repeat it at several sizes and use conservative data. For debt, model how collateral value and borrow value change together. Include accrued interest and liquidation penalties where documented by the protocol.
FINRA identifies liquidity risk as the possibility that an investment cannot be sold quickly or efficiently. Crypto adds around-the-clock markets but not guaranteed depth; a market can be open and still be functionally illiquid.
## T — Thesis and time horizon [#t--thesis-and-time-horizon]
Every material position should have a current reason for existing in the portfolio. Record:
* original thesis and date;
* expected mechanism of value or utility;
* time horizon;
* evidence that would strengthen the thesis;
* evidence that would falsify it;
* known catalysts and risks;
* what has changed since entry;
* whether the position's current weight reflects intention or price drift.
Avoid retrofitting a new thesis after the old one fails. “Short-term trade became long-term conviction” is a common form of thesis drift.
When a thesis needs a full reset, rerun the same [pre-buy due diligence process](/blog/what-to-check-before-buying-crypto) against current evidence instead of editing only the conclusion.
Investor.gov notes that faster growth in one holding can move a portfolio away from its intended allocation and change its risk level. Whether to rebalance is a personal decision with tax, fee, and liquidity consequences. The health check should identify drift, not issue a trade instruction.
### Thesis-age formula [#thesis-age-formula]
Use `thesis age = current date - last evidence-based review date`.
Do not set one universal maximum. A protocol upgrade thesis may need review around governance and release events; a stable custody thesis may use a slower cadence but require immediate review after an incident. The formula makes staleness visible.
## H — Habits, alerts, and review cadence [#h--habits-alerts-and-review-cadence]
A healthy portfolio process surfaces material changes without producing so much noise that every alert is ignored.
Map each thesis risk to an observable signal:
| Risk | Possible signal | Verification source |
| -------------------- | ------------------------------------------------------- | ---------------------------------------- |
| Stablecoin depeg | Price deviation plus redemption/reserve news | Issuer, venues, primary disclosures |
| Protocol exploit | Official incident notice and abnormal contract flows | Protocol, explorer, security researchers |
| Insider unlock | Vesting event and receiving-wallet activity | On-chain vesting contract |
| Liquidity rug | LP-position movement or large withdrawal | Pool and locker contracts |
| Exchange exposure | Withdrawal status, incident, or solvency disclosure | Exchange and regulators |
| Thesis deterioration | Usage, fees, governance, roadmap, or competitive change | Protocol and primary data |
Price alone is a lagging summary of many causes. Use [Crypto Alerts Beyond Price](/blog/crypto-alerts-beyond-price) to design a layered alert system.
BlockMind's agent runs periodic monitoring and can accept durable alert instructions in plain language. It is not a tick-level execution service; alerts can take up to a few hours and inform rather than act.
## The health-check evidence table [#the-health-check-evidence-table]
Create one table per review:
| Finding | Evidence | Impacted value | Confidence | Owner/action | Recheck trigger |
| ------------------------ | --------------------------- | -------------------: | ---------- | ------------------------------ | --------------------- |
| Missing staking position | Address + protocol position | Hypothetical $5,000 | High | Add to analytical view | Position changes |
| Hidden ETH look-through | LST + collateral mapping | Hypothetical $28,000 | Medium | Review exposure | Protocol/price change |
| Old exchange API key | Exchange key list | Privacy exposure | High | Rotate/revoke after validation | Quarterly review |
| Shallow token depth | Named venue depth snapshot | Hypothetical $30,000 | Medium | Document constraint | Depth changes 25% |
| Thesis not reviewed | Notebook date | Hypothetical $12,000 | High | Re-run primary-source review | Governance event |
“Action” can mean gather evidence, correct data, rotate a credential, or consult a professional. It does not have to mean trade.
## A 30-minute minimum viable review [#a-30-minute-minimum-viable-review]
If time is limited:
1. Reconcile every account and wallet.
2. Calculate top-one and top-three exposure.
3. Look through wrappers, collateral, and ecosystem duplicates.
4. List every trade- or withdrawal-enabled API key.
5. Test liquidity for the three largest non-major positions.
6. Read the last thesis note for the five largest exposures.
7. Confirm alerts exist for custody, protocol, liquidity, and unlock risks.
8. Save the timestamp and unresolved data gaps.
This is a triage pass, not a complete review.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Portfolio trackers may omit transactions, protocols, chains, or liabilities.
* Public prices can diverge from executable prices.
* Correlations change, especially during stress.
* Wallet labels and protocol classifications can be wrong.
* Diversification reduces some risks but cannot remove market-wide loss.
* A low concentration measure does not protect against a shared bridge, oracle, custodian, or stablecoin dependency.
* Security controls reduce risk but cannot guarantee recovery from fraud, compromise, or insolvency.
* Personal tax, legal, cash-flow, and suitability questions require qualified advice.
## How BlockMind supports a health check [#how-blockmind-supports-a-health-check]
Once portfolios are connected for balance-only use, ask your agent to map holdings, allocation, ecosystem exposure, performance drivers, and position changes. It can preserve review notes in the [Notebook](/docs/memory/notebook) and use the portfolio in the Morning Brief and monitoring passes.
The agent's output can be incomplete or wrong. Reconcile it with source accounts and primary protocol data. BlockMind never trades, moves funds, or tells you what to buy or sell.
## The Bottom Line [#the-bottom-line]
A crypto portfolio health check is a data, risk, and process audit. HEALTH makes the output explicit: holdings integrity, exposure, access, liquidity, thesis, and habits.
Do not optimize the dashboard before validating the data. Do not celebrate diversification before looking through wrappers and shared dependencies. And do not rely on price alerts when the real failure may begin in a contract, bridge, custodian, or thesis.
This article is for research and education, not financial, legal, tax, or security advice.
## Sources [#sources]
1. [Investor.gov: Asset allocation, diversification, and rebalancing](https://www.investor.gov/introduction-investing/getting-started/asset-allocation)
2. [FINRA: Concentration risk](https://www.finra.org/investors/insights/concentration-risk)
3. [FINRA: Investment risk and liquidity risk](https://www.finra.org/investors/investing/investing-basics/risk)
4. [CFTC: Risks of virtual-currency trading](https://www.cftc.gov/sites/default/files/2019-12/customeradvisory_urvct121517.pdf)
5. [BlockMind: Portfolio data and connection methods](/docs/data/portfolios)
6. [BlockMind: Portfolio analysis](/docs/capabilities/portfolio-analysis)
7. [BlockMind: Trust and security](/docs/security)
---
# Crypto Pump-and-Dump Signs: How to Separate Momentum From Manipulation
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: problem-solving
Primary keyword: crypto pump and dump signs
Keywords: crypto pump and dump signs, how to spot crypto pump and dump, crypto market manipulation, wash trading crypto, token pump warning signs
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** The strongest crypto pump-and-dump warning is not a price spike alone. It is a *combination* of thin liquidity, concentrated or linked ownership, unverifiable promotion, artificial-looking volume, urgent coordination, and insider distribution into new demand. These signals justify deeper investigation; they do not prove intent by themselves.
***
A legitimate catalyst can produce a fast rally. So can coordinated manipulation. A chart cannot tell you which story is true without market, on-chain, and off-chain context.
This is one signal family inside the broader [crypto rug pull checker workflow](/blog/crypto-rug-pull-checker-guide), which explains how scanners, contract controls, liquidity, holders, and off-chain claims fit together.
The [CFTC's pump-and-dump advisory](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html) warns specifically about thinly traded or new tokens promoted through social media, messaging groups, false news, urgency, and sudden price spikes. Its advice is simple: do not buy because of a single tip or spike, and research the asset and people behind it.
## What is a crypto pump-and-dump? [#what-is-a-crypto-pump-and-dump]
A crypto pump-and-dump is a scheme in which organizers accumulate or control an asset, create misleading demand or attention, and sell into the resulting price increase, leaving later buyers with losses.
The phases often look like this:
1. **Positioning:** organizers create or quietly accumulate a thinly traded token.
2. **Ignition:** coordinated promotion, false claims, bots, wash trading, or countdown groups attract attention.
3. **Distribution:** insiders sell into new demand while public messaging remains bullish.
4. **Collapse:** demand disappears, liquidity is removed or overwhelmed, and late holders cannot exit near the displayed price.
The CFTC's enforcement case against John McAfee and Jimmy Watson alleged this sequence: secret accumulation, misleading social promotion, then sales into the resulting price increase. Allegations and case outcomes depend on their records, but the sequence illustrates why undisclosed promoter ownership matters.
## The eight most useful warning signs [#the-eight-most-useful-warning-signs]
### 1. A price spike without a verifiable catalyst [#1-a-price-spike-without-a-verifiable-catalyst]
Look for a dated, primary source: protocol release, governance result, filing, exchange notice, exploit resolution, or other concrete event. A recycled roadmap item, anonymous screenshot, or influencer hint is not equivalent evidence.
Absence of a catalyst does not prove manipulation. Markets move on positioning and information that may not yet be public. It does increase the need to inspect who bought, where volume came from, and whether promotion preceded the move.
### 2. Thin executable liquidity [#2-thin-executable-liquidity]
Market capitalization is not exit capacity. A token can show a large paper value while only a small pool supports trading. That makes it easier for modest buying to move price and harder for later holders to sell.
Check pool depth around several order sizes, not just headline liquidity. Our [liquidity locks guide](/blog/liquidity-locks-crypto-explained) explains ownership and removal risk; a lock does not create depth.
### 3. Volume that does not fit the market [#3-volume-that-does-not-fit-the-market]
High reported volume alongside shallow liquidity, repetitive trade sizes, rapid back-and-forth activity, or a small set of funded wallets can indicate wash trading.
The SEC's [2024 market-manipulation charges](https://www.sec.gov/newsroom/press-releases/2024-166) alleged that promoters and purported market makers used self-trading and bots to create artificial volume. The allegations demonstrate a core lesson: volume can be manufactured and should not be treated as independent demand without scrutiny.
Chainalysis similarly emphasizes that its on-chain heuristics detect patterns, not intent. Arbitrage and MEV can resemble some suspicious behavior, so off-chain evidence and wallet relationships matter.
### 4. Countdown groups and forced urgency [#4-countdown-groups-and-forced-urgency]
“Announcement in ten minutes,” “everyone buy at once,” “do not sell,” and “share this everywhere” are classic coordination signals. Organizers know the asset before the crowd and can position first.
The CFTC warns that messaging apps can add users to fraudulent groups and that pump organizers may coordinate anonymous participants. A large group count does not mean independent demand; accounts can be fake, compromised, or controlled.
### 5. Unverifiable partnerships or news [#5-unverifiable-partnerships-or-news]
Manipulators may circulate a fake listing, partnership, celebrity endorsement, or media page. Verify it on the exchange, partner, regulator, or person's official domain—not through the token's repost.
Be careful with screenshots. Inspect the URL and publication date. AI-generated media and hacked accounts make visual polish weak evidence.
### 6. Concentrated or linked holders [#6-concentrated-or-linked-holders]
If the deployer, promoter, market maker, or related wallets control a large liquid share, they can supply the dump. Label pools, bridges, treasuries, vesting contracts, burn addresses, and exchanges before interpreting the holder list.
Then trace funding and transfers among unlabeled top wallets. One actor can split holdings across many addresses. Wallet count is not beneficial-owner count.
### 7. Promoters hide compensation or positions [#7-promoters-hide-compensation-or-positions]
A promoter who owns the asset has a conflict. Ask whether they disclose holdings, compensation, lockups, and sale history. Disclosure does not make a promotion sound, but concealment removes information needed to judge incentives.
### 8. Insider outflows during relentless bullish messaging [#8-insider-outflows-during-relentless-bullish-messaging]
Distribution can appear as transfers to exchange-associated addresses, sales through pools, liquidity withdrawal, or movement through intermediary wallets. An exchange deposit is not proof of sale; it is a reason to inspect subsequent transactions and alternative explanations.
The strongest pattern is contradiction: public claims say “we are not selling” while verifiably linked wallets reduce exposure.
## The SCOPE evidence matrix [#the-scope-evidence-matrix]
Use five evidence categories: **Social, Catalyst, Ownership, Pool, Execution**.
| Category | Questions | Stronger evidence | Weak evidence |
| ------------- | ---------------------------------------------------- | ---------------------------------------------------------- | ------------------------------------------------- |
| **Social** | Who started the narrative? Is promotion coordinated? | Dated posts, disclosure, account history | Follower count, screenshots, anonymous repetition |
| **Catalyst** | What changed in the underlying project? | Official release, filing, governance record, deployed code | Rumor, teaser, “big news soon” |
| **Ownership** | Who can supply the market? | Labeled holder map, funding links, vesting contracts | Raw holder count |
| **Pool** | Can real positions exit? | Depth by size, pool ownership, lock record | Market cap and daily volume alone |
| **Execution** | Are trades economically independent? | Address-level flow and venue comparison | Candlestick shape alone |
Do not convert the matrix into a criminal accusation. Use it to state what is observed and what remains unknown.
## A volume-to-depth stress test [#a-volume-to-depth-stress-test]
One useful screening ratio is:
`reported 24-hour volume ÷ executable two-sided depth within a chosen price band`.
**Hypothetical example:** A token reports $8 million in daily volume, but an order-book or pool-depth tool shows only $80,000 of combined depth within 5% of the current price. The ratio is `8,000,000 ÷ 80,000 = 100`.
A ratio of 100 does not prove wash trading. Active turnover can legitimately exceed standing depth, and data sources define depth differently. It does show that the headline volume and near-price liquidity tell very different stories. Investigate venue concentration, repetitive trades, address funding, and whether reported volume can be independently reproduced.
## Legitimate momentum vs suspected manipulation [#legitimate-momentum-vs-suspected-manipulation]
| Observation | Legitimate explanation to test | Manipulation explanation to test |
| ---------------------- | --------------------------------------- | -------------------------------------- |
| Fast price increase | Material product or market catalyst | Coordinated buying into thin liquidity |
| Social mentions surge | Organic response to news | Bots, paid promotion, countdown group |
| Volume rises | New venue and independent buyers | Wash trading or self-trading |
| Large wallet transfers | Custody change, market-making inventory | Insider distribution |
| Liquidity changes | Planned migration or LP rebalancing | Exit liquidity removed |
Good analysis carries both columns until evidence rules one out.
## What to do when you suspect a pump [#what-to-do-when-you-suspect-a-pump]
* Do not join a coordinated pump group or forward its instructions.
* Preserve URLs, messages, timestamps, transaction hashes, token and pool addresses.
* Verify the contract address and current sellability without risking funds; see [What Is a Honeypot Token?](/blog/honeypot-crypto-token).
* Map holder and liquidity control.
* Check primary sources for every claimed catalyst.
* Reconcile reported volume across venues and analytics sources.
* If you believe fraud occurred, use the reporting channels of the relevant exchange and authorities in your jurisdiction.
* Ignore strangers offering fund recovery for an upfront crypto payment or seed phrase.
## A pre-purchase pump-risk checklist [#a-pre-purchase-pump-risk-checklist]
* [ ] Exact token and pool addresses verified
* [ ] Catalyst found on a primary source
* [ ] Promoters disclose compensation and holdings
* [ ] Top-holder map adjusted for system addresses
* [ ] Deployer-linked wallets traced
* [ ] Liquidity depth tested at realistic sizes
* [ ] Liquidity ownership and unlocks checked
* [ ] Volume compared across independent venues
* [ ] Repetitive or self-funded trade patterns examined
* [ ] Contract checked for sell restrictions and mutable taxes
* [ ] Counterevidence documented
* [ ] No action based solely on urgency or social proof
For a broader token review, use the [memecoin due-diligence checklist](/blog/memecoin-due-diligence-checklist) or [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps).
## Limitations and counterevidence [#limitations-and-counterevidence]
No public checklist can establish a person's intent. Wallet labels can be wrong, exchange flows can have innocent explanations, and legitimate markets can become highly reflexive. A pump-and-dump is a factual and often legal conclusion requiring evidence beyond an unusual chart.
There is also survivorship bias: spectacular legitimate rallies are remembered, while many ordinary rallies fade without fraud. “Price fell after hype” is not enough. State observations narrowly—“volume appears concentrated,” “wallets share a funding source,” or “the partnership is not confirmed”—and separate them from inference.
## How BlockMind can help [#how-blockmind-can-help]
A BlockMind agent can help gather primary sources, map public market and on-chain evidence, save the thesis in a [Notebook](/docs/memory/notebook), and monitor holdings or tracked assets periodically. It cannot prove intent, identify every wallet owner, or guarantee timely detection. Monitoring is periodic rather than tick-level. BlockMind never trades or tells you what to buy or sell.
## The Bottom Line [#the-bottom-line]
Do not diagnose a pump-and-dump from a candle. Look for convergence across SCOPE: coordinated social promotion, an unverifiable catalyst, concentrated ownership, shallow pools, and trading that may not represent independent demand.
One red flag is a question. Several mutually reinforcing red flags are a reason to stop and investigate. Neither is permission to make unsupported accusations.
This article is for research and education, not financial advice or legal advice.
## Sources [#sources]
1. [CFTC: Beware virtual-currency pump-and-dump schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
2. [CFTC: Digital-asset fraud patterns](https://www.cftc.gov/LearnAndProtect/digitalassetfrauds)
3. [CFTC: Enforcement action describing alleged secret accumulation and promotion](https://www.cftc.gov/PressRoom/PressReleases/8366-21)
4. [SEC: 2024 charges concerning alleged artificial crypto trading volume](https://www.sec.gov/newsroom/press-releases/2024-166)
5. [U.S. Department of Justice: Operation Token Mirrors](https://www.justice.gov/usao-ma/pr/eighteen-individuals-and-entities-charged-international-operation-targeting-widespread)
6. [Chainalysis: 2025 market-manipulation methodology](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/)
---
# Crypto Research Workflow Template: A Copyable Evidence Ledger
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: problem-solving
Primary keyword: crypto research workflow template
Keywords: crypto research workflow template, crypto research template, cryptocurrency analysis template, DYOR template, crypto evidence ledger
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A good crypto research workflow leaves an audit trail. Define the asset and question, inventory claims, collect primary evidence, align dates and definitions, run market, protocol, security, token, and liquidity checks, record contradictory evidence, then set explicit invalidation and review triggers. The copyable template below is the working document; it does not produce a buy or sell instruction.
***
## What is a crypto research workflow? [#what-is-a-crypto-research-workflow]
A crypto research workflow is a repeatable sequence for turning an asset or protocol question into a sourced, testable, updateable research record.
It should answer:
* What exactly did you research?
* Which claim or decision context set the scope?
* What evidence supports and contradicts each material conclusion?
* Which calculations can another reader reproduce?
* What remains unknown?
* What event makes the report stale?
This page owns the working document and evidence ledger. Use [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research) to decide which tasks your agent can gather and which judgments must remain human. Use [What to Check Before Buying Crypto](/blog/what-to-check-before-buying-crypto) for the concise evidence categories and the [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps) for the exhaustive checklist.
**Scope and assumptions:** The template covers public research on fungible cryptoassets and protocols. Remove irrelevant modules rather than filling them with guesses. It does not determine personal suitability, position size, legal status, or tax treatment.
*This is research, not financial advice. BlockMind does not tell you what to buy or sell, and its agent cannot trade, withdraw, approve, or move funds.*
## How to use this template [#how-to-use-this-template]
1. Copy the master template into a durable note.
2. Complete the scope before gathering evidence.
3. Give every material claim a stable ID.
4. Prefer primary sources and preserve retrieval dates.
5. Recalculate important formulas.
6. Keep observations separate from interpretations.
7. Require counterevidence and unknowns.
8. End with review triggers, not a permanent verdict.
Do not delete old conclusions when evidence changes. Append a dated revision so the reasoning remains inspectable.
## The master crypto research template [#the-master-crypto-research-template]
Copy everything inside this block:
```text
# [ASSET / PROTOCOL] RESEARCH RECORD
STATUS: Draft / Reviewed / Needs update / Archived
CREATED (UTC):
LAST VERIFIED (UTC):
OWNER:
REVIEWER:
NEXT CALENDAR REVIEW:
## 1. RESEARCH BRIEF
Exact asset:
- Name:
- Symbol:
- Chain/network:
- Contract or native identifier:
- Canonical website:
- Canonical repositories:
Question:
Time horizon:
Decision context:
Comparison set:
Out of scope:
Working hypothesis:
Evidence that would support it:
Evidence that would contradict it:
## 2. SOURCE AND DEFINITION RULES
Research timestamp:
Timezone:
Reference currency:
Price methodology:
Circulating-supply definition:
Volume methodology:
On-chain label provider and caveats:
Source priority:
1. Deployed state / signed primary record / official registry
2. Current official documentation or governance
3. Independent auditor, regulator, or methodology-led data provider
4. Reputable secondary analysis
5. Social commentary as a lead only
## 3. CLAIM LEDGER
| ID | Material claim | Claim owner | Evidence required | Status | Contradiction | Freshness |
|----|----------------|-------------|-------------------|--------|---------------|-----------|
| C-01 | | | | | | |
Statuses: verified / partly verified / unverified / contradicted / stale
## 4. PROJECT AND RIGHTS
Problem and user:
Mechanism in plain language:
Tokenholder rights:
Where rights are enforced:
Legal entity / governance body:
Key dependencies and trust assumptions:
Actor and permission map:
| Actor | Power | Constraint | Failure mode | Evidence |
|-------|-------|------------|--------------|----------|
## 5. PRODUCT AND ADOPTION
Current shipped product:
Target user:
Usage metrics and definitions:
Retention evidence:
Fees / revenue and definitions:
Incentive-adjusted interpretation:
Competitors and switching costs:
What product success does NOT prove about the token:
## 6. TOKEN AND SUPPLY
Reference price and timestamp:
Circulating supply:
Unlocked supply:
Total supply:
Maximum supply / change authority:
Market cap calculation:
FDV calculation:
Allocation table:
| Class | Tokens/% | Recipient | Lock/vesting | Control | Source |
|-------|----------|-----------|--------------|---------|--------|
3/6/12/24-month supply bridge:
| Window | Unlocks | Emissions | Burns | Estimated new circulation | Assumptions |
|--------|---------|-----------|-------|---------------------------|-------------|
Token utility and value-capture chain:
Supply discrepancies:
## 7. TEAM, GOVERNANCE, AND DEVELOPMENT
Team identity evidence:
Prior work and counterparty confirmation:
Conflicts and compensation:
Governance:
- Proposal threshold:
- Quorum/approval:
- Delegation concentration:
- Guardian/veto/admin powers:
- Execution history:
Repository-to-product map:
| Component | Repository | Review/test evidence | Release | Live version/address |
|-----------|------------|----------------------|---------|----------------------|
Maintenance concentration and security process:
## 8. SECURITY
Current deployments:
Proxy / implementation / admin:
Pause, freeze, mint, rescue, and upgrade powers:
Audit map:
| Auditor | Final report | Commit/scope | Findings unresolved | Current deployment match |
|---------|--------------|--------------|---------------------|--------------------------|
Bug bounty:
Testing / formal methods:
Dependencies / oracles / bridges:
Incident history and response:
Residual risks:
## 9. MARKET AND LIQUIDITY
Market regime:
Higher-timeframe structure:
Relative strength:
Spot participation:
Leverage/funding/open interest:
Liquidity tests (timestamped):
| Venue/route | Spread | 1% bid/ask depth | Small impact | Medium impact | Stress impact |
|-------------|--------|------------------|--------------|---------------|---------------|
Venue and LP concentration:
Quote-asset and access risks:
## 10. ON-CHAIN AND COMMUNITY
Holder/entity concentration:
Treasury and known wallets:
Exchange/whale flows with label confidence:
Internal-flow adjustments:
Community sample windows:
Support resolution:
Critical-question handling:
Contribution/governance output:
Incentive or coordinated-looking activity:
## 11. CATALYSTS AND FAILURE MODES
| Date/window | Event | Mechanism | Confirmation | Counterevidence | Source |
|-------------|-------|-----------|--------------|-----------------|--------|
Pre-mortem: Assume the thesis fails. Most plausible causes:
1.
2.
3.
## 12. EVIDENCE SUMMARY
Strongest supporting evidence:
1.
2.
3.
Strongest contradictory evidence:
1.
2.
3.
Unknowns that could change the conclusion:
1.
2.
3.
## 13. CONCLUSION
Observation summary:
Interpretation:
Confidence: low / medium / high, with reason
Structural invalidation:
What this report does NOT conclude:
## 14. MONITORING AND CHANGE LOG
Event triggers:
- Contract upgrade or admin change
- Material audit or incident
- Supply/unlock threshold
- Liquidity deterioration threshold
- Governance or team change
- Product/adoption threshold
- Thesis invalidation
Change log:
| Date | Claim/section | Old evidence | New evidence | Conclusion changed? | Reviewer |
|------|---------------|--------------|--------------|---------------------|----------|
## 15. SOURCE LEDGER
| ID | Source | Type | Published | Retrieved UTC | Supports | Archive/hash | Limits |
|----|--------|------|-----------|---------------|----------|--------------|--------|
```
## The minimum viable workflow [#the-minimum-viable-workflow]
When a full report is not justified, do not fake completeness. Use this 20-minute triage record:
```text
Exact asset/contract:
Question and horizon:
1. Current product and token right:
2. Circulating/total/max supply definitions:
3. Next material unlock or emission:
4. Current deployment and privileged controls:
5. Audit-to-deployment match:
6. Liquidity at one relevant size across two routes:
7. Team/repository identity:
8. Strongest supporting evidence:
9. Strongest contradiction:
10. Unknown that blocks a conclusion:
Decision: stop / commission full review
Sources and retrieval times:
```
Triage should decide whether deeper research is worthwhile. It should not manufacture conviction from a shallow pass.
## The source ledger: how to make evidence reusable [#the-source-ledger-how-to-make-evidence-reusable]
Every source should answer four questions:
1. **Provenance:** Who published or controls it?
2. **Scope:** Which asset, component, date, and definition does it cover?
3. **Support:** Which exact claim does it substantiate?
4. **Limits:** What can it not establish?
Example:
| ID | Source | Type | Supports | Limit |
| ---- | --------------------------- | ------------------------- | ------------------------------ | ------------------------------------- |
| S-01 | Verified vesting contract | Deployed primary evidence | Releasable amount at timestamp | Does not prove sale intent |
| S-02 | Auditor-hosted final report | Independent scoped review | Findings for commit `abc` | Does not cover later deployment |
| S-03 | Project usage dashboard | Issuer data | Displayed daily users | Methodology and bot filtering unclear |
The CFTC recommends extensive research into token affiliates, rights, and how funds will be used ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)). A ledger turns that broad duty into traceable work.
## The calculation sheet [#the-calculation-sheet]
Keep formulas next to inputs.
### Valuation [#valuation]
```text
Market cap = price × circulating supply
FDV = price × maximum supply
Circulation ratio = circulating supply / maximum supply
```
Use [Market Cap vs FDV](/blog/market-cap-vs-fdv-crypto) for definitions and scenario limits.
### Unlock impact [#unlock-impact]
```text
Unlock ratio = scheduled tokens / current circulating supply
Net scheduled change = unlocks + emissions − burns
```
Separate scheduled, releasable, claimed, transferable, and circulating amounts.
### Liquidity [#liquidity]
```text
Spread % = (ask − bid) / midpoint × 100%
Sell impact % = (midpoint − average simulated execution) / midpoint × 100%
```
### Concentration [#concentration]
```text
Top-N share = balances controlled by top N relevant entities / eligible supply
```
Exclude burn, bridge, exchange, pool, staking, and protocol contracts only when the research question and methodology justify exclusion.
## A worked hypothetical workflow excerpt [#a-worked-hypothetical-workflow-excerpt]
**This is a fictional example. It is not a real asset analysis or recommendation.**
Question: Does “Meridian Protocol” have independently verifiable product traction that could create token demand over 12 months?
Claim ledger excerpt:
| ID | Claim | Evidence | Status | Contradiction |
| ---- | ------------------------- | ------------------------ | --------------- | ---------------------------------------------------------- |
| C-01 | 40,000 monthly users | Issuer dashboard only | Unverified | On-chain unique callers materially lower |
| C-02 | Token required for fees | Current docs + contracts | Verified | Fees can also be paid through sponsored transactions |
| C-03 | Fixed supply | Verified token code | Partly verified | Governance can replace token contract through upgrade path |
| C-04 | Current contracts audited | Auditor report | Contradicted | Report covers prior implementation |
Evidence summary:
* Support: actual product, growing transaction count, fee token role
* Counterevidence: user methodology opaque, sponsored usage weakens direct token demand, audit mismatch
* Unknown: retention after incentives end
Conclusion:
> Meridian has verifiable usage and a token role, but the strength of value capture is uncertain because sponsored transactions reduce mandatory demand. User count and current audit coverage remain unresolved. Review after the incentive program ends and after a current-implementation audit.
The workflow does not force a yes/no outcome when evidence is mixed.
## How AI should fit into the workflow [#how-ai-should-fit-into-the-workflow]
Your agent is useful for:
* Gathering and normalizing public sources
* Extracting claims from long documents
* Recalculating transparent formulas
* Comparing current and older pages
* Monitoring named triggers
* Formatting the ledger and change log
Keep human control over:
* Scope and materiality
* Source independence
* Interpretation of incentives and people
* Risk tolerance and personal decisions
* Final verification of decisive evidence
NIST’s AI Risk Management Framework emphasizes governance, mapping context, measurement, and management, while its AI Resource Center highlights testing, evaluation, verification, and validation ([NIST AIRC](https://airc.nist.gov/)). Those concepts translate well here: define the context, test the output, and preserve accountability.
Use [How to Verify AI Crypto Analysis](/blog/how-to-verify-ai-crypto-analysis) before moving any AI-generated sentence into the verified column.
## Workflow quality checks [#workflow-quality-checks]
Before marking a report reviewed:
* The exact asset and current contract are unambiguous
* Every material conclusion maps to claim IDs and sources
* Dates, units, and definitions align
* Important formulas were recalculated
* Current deployments map to code and audit scope
* Liquidity tests state size, pair, venue, and time
* Transfers are not described as intent without evidence
* Strongest counterevidence is present
* Unknowns are visible
* Review dates and event triggers are set
* A second reader can reproduce the decisive steps
## Limitations and counterevidence [#limitations-and-counterevidence]
* A complete template can encourage box-checking; remove irrelevant sections and deepen material ones.
* Public evidence may be incomplete or revised.
* The workflow cannot guarantee safety or predict price.
* More data can increase noise if the question is poorly scoped.
* AI can reproduce source errors or generate plausible but unsupported claims.
* A reviewer can share the original author’s bias.
The workflow improves traceability, not certainty.
## Using the template in BlockMind [#using-the-template-in-blockmind]
Save the copied template in your [Notebook](/docs/memory/notebook) so it receives a stable NB-number and can be updated over time. Ask your BlockMind agent to populate the mechanical fields with current sources, preserve unknowns, and create monitoring conditions for material triggers. Its [research capabilities](/docs/capabilities/research), [on-chain intelligence](/docs/capabilities/on-chain), and [browser](/docs/capabilities/browser) can support different evidence layers.
Example request:
> “Create a research record for \[asset] using this template. Complete the source and claim ledgers first, stop any claim from becoming verified without direct evidence, show formulas and timestamps, and ask me to judge the final interpretation.”
## The Bottom Line [#the-bottom-line]
The best crypto research template is not the one with the most fields. It is the one that makes weak evidence, conflicting definitions, calculations, counterevidence, and staleness impossible to hide.
Copy the workflow, scope the question, preserve the claim and source ledgers, separate observation from interpretation, and set the events that reopen the work. Research earns trust when another reader can reproduce it and when future evidence can change it cleanly.
## Sources [#sources]
* [CFTC — Use Caution When Buying Digital Coins or Tokens](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)
* [NIST — AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework)
* [NIST — AI Resource Center](https://airc.nist.gov/)
* [Ethereum.org — Smart Contract Security and Audit Limits](https://ethereum.org/developers/docs/smart-contracts/security)
* [CoinMarketCap — Market Capitalization Methodology](https://support.coinmarketcap.com/hc/en-us/articles/360043836811-Market-Capitalization-Cryptoasset-Aggregate)
* [EUR-Lex — Markets in Crypto-Assets Disclosure Categories](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32023R1114)
---
# Crypto Rug Pull Checker Guide: What Scanners Can and Cannot Prove
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: problem-solving
Primary keyword: crypto rug pull checker
Keywords: crypto rug pull checker, rug pull scanner, token scam checker, how to check a crypto token, rug pull detection
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A crypto rug pull checker can flag known contract patterns, failed sell simulations, dangerous permissions, holder concentration, or removable liquidity. It cannot prove a token is safe, identify every coordinated wallet, predict future admin actions, or verify a team's off-chain claims. Use several tools, verify the exact contract address, and keep an evidence ledger.
***
If a scanner shows “low risk,” the correct interpretation is: **the checks it ran did not detect the risks they cover at that moment**. That is very different from “this token cannot rug.”
Rug pulls can happen through several mechanisms: removing liquidity, making sales impossible, changing taxes, minting supply, dumping concentrated holdings, compromising an admin key, or simply abandoning a project after raising attention. No single API observes every mechanism.
Chainalysis makes the same methodological distinction in its [2025 market-manipulation research](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/): on-chain behavior can identify suspicious patterns, but behavior alone does not prove intent. Its criteria are useful starting points for investigation, not verdicts about every token.
## What a rug pull checker actually checks [#what-a-rug-pull-checker-actually-checks]
Different tools cover different slices of risk:
| Tool type | Useful for | Cannot establish by itself |
| ------------------------ | ------------------------------------------------------------------------------------ | ------------------------------------------------------------------ |
| Swap simulator | Whether a modeled buy and sell succeeds; estimated tax and gas | That future sells will work or every route behaves the same |
| Contract-pattern scanner | Known malicious bytecode/source patterns and risky functions | That novel or obfuscated logic is harmless |
| Token-security API | Permissions, taxes, blacklist/whitelist logic, mintability, proxy and holder signals | Team identity, honest governance, or off-chain promises |
| Block explorer | Verified source, contract creator, holders, transactions, admin calls | Beneficial ownership behind every address |
| Liquidity-lock record | Amount, owner, pool, unlock time, and locker contract | Token contract safety, team-token vesting, or permanent liquidity |
| Audit report | Findings in a defined code version and scope | Safety of excluded code, later upgrades, operations, or governance |
[Honeypot.is](https://docs.honeypot.is/ishoneypot) documents a simulation response that may include buy tax, sell tax, transfer tax, gas, maximum buy/sell amounts, and a honeypot result. Its documentation also says fields may be absent if a simulation fails or a result cannot be determined. “Unknown” is not “safe.”
[GoPlus](https://docs.gopluslabs.io/reference/response-details) returns token-security fields including a honeypot indicator, but its response documentation explicitly allows no value when the result is unknown. [Token Sniffer](https://tokensniffer.readme.io/reference/introduction) analyzes source code and bytecode against known scam patterns and can add holder and liquidity metrics. Each is useful; each has a bounded detection surface.
## Before any check: verify the contract address [#before-any-check-verify-the-contract-address]
Token names and tickers are not identifiers. Anyone can deploy a token with a familiar name or symbol. Start from the project's official documentation, then cross-check the address against at least one independent authoritative listing or protocol interface.
Confirm:
* chain and network;
* full contract or mint address, not a shortened visual match;
* official website and documentation domain;
* trading pair address;
* whether the contract is a proxy and, if so, its implementation;
* whether the source code is verified.
Etherscan explains that [source verification](https://info.etherscan.com/how-to-verify-contracts/) matches submitted source code to deployed bytecode and publishes the source for inspection. Verification improves transparency. It is not an audit and does not certify that the logic is safe.
If the address cannot be established confidently, stop. A sophisticated scan of the wrong token produces confidently irrelevant output.
## The CLEAR workflow [#the-clear-workflow]
Use five layers instead of one score: **Contract, Liquidity, Equity, Accountability, Reality**.
### C — Contract behavior and control [#c--contract-behavior-and-control]
Run at least one swap simulation and one contract/security scan. Then inspect the explorer.
Look for:
* successful buy *and* sell simulation;
* taxes and whether an admin can change them;
* maximum transaction or wallet limits;
* blacklist, whitelist, pause, cooldown, and transfer-gating logic;
* mint, burn-from, confiscation, or balance-modification authority;
* proxy upgradeability and the address controlling upgrades;
* verified source and correspondence to the current implementation;
* privileged roles, not just a single `owner` field.
Do not treat “ownership renounced” as a universal guarantee. OpenZeppelin's [access-control documentation](https://docs.openzeppelin.com/contracts/5.x/access-control) shows that contracts can use multiple roles, another contract as owner, or delayed administration. Renouncing one owner does not prove every privileged role disappeared.
**Stop rule:** A failed or indeterminate sell simulation, unexplained blacklist logic, arbitrary minting, or an upgrade controller the team will not explain is enough to pause the investigation.
### L — Liquidity control and exit conditions [#l--liquidity-control-and-exit-conditions]
Find the real trading pool and identify who owns the position representing liquidity. On Uniswap, [v2 positions use fungible UNI-V2 tokens while v3 and v4 positions use NFTs](https://support.uniswap.org/hc/en-us/articles/20980786685069-Why-is-liquidity-position-ownership-represented-by-tokens-or-NFTs). The wallet holding that token or NFT controls the position unless it is locked or transferred.
Record:
* pool address and DEX version;
* current liquidity and depth around realistic trade sizes;
* each liquidity owner's share;
* locker contract address, not just a badge on the project's site;
* percentage locked and exact unlock timestamp;
* whether locks cover the active pool and price range;
* other pools from which liquidity can migrate or be removed;
* the exit impact of a realistic position.
Read [Liquidity Locks Explained](/blog/liquidity-locks-crypto-explained) before treating a lock as dispositive. A partial, short, or wrong-pool lock can look reassuring while leaving the meaningful liquidity removable.
### E — Equity and holder concentration [#e--equity-and-holder-concentration]
Review holders and label obvious system addresses: burn addresses, pool contracts, bridges, vesting contracts, treasuries, and centralized exchanges. Then examine the remaining wallets.
Questions:
* What share do the deployer and linked wallets control?
* Were top wallets funded by the same source?
* Are tokens split across many fresh addresses?
* Do insiders have documented vesting, and does it match on-chain reality?
* Can a small number of holders overwhelm available liquidity?
* Are reported “holders” largely dust or airdropped wallets?
Wallet count is easy to inflate. Chainalysis found that one controller can manage many addresses in suspicious wash-trading patterns. Address count is not the same as independent ownership.
### A — Accountability and evidence [#a--accountability-and-evidence]
Move off-chain. Verify whether the team, audit, legal entity, roadmap, partnerships, and repositories exist independently of the project's own marketing.
* Open the audit on the auditor's domain.
* Match audited commit, chain, and contract address to production.
* Check unresolved findings and excluded components.
* Verify named partners on the partner's own site.
* Inspect repository history for substantive work, not just a recent code dump.
* Distinguish pseudonymity from fabricated identity.
* Archive important claims with dates.
An audit is bounded evidence. It does not cover economic manipulation, compromised keys, undisclosed linked wallets, or code deployed after the report.
### R — Reality: product, market, and contradictions [#r--reality-product-market-and-contradictions]
Finally, test whether the observable project matches the story.
* Is there a working product or only a token and roadmap?
* Do usage, fees, liquidity, and developer activity support claimed traction?
* Is social attention driven by analysis or referral spam and price promises?
* Does reported volume align with liquidity and independent venues?
* What evidence would falsify the bull story?
* What changed since the scanner's timestamp?
The CFTC advises people not to buy tokens based on a single social-media tip or sudden price spike and to separate hype from facts. That applies even when a clean scanner screenshot is part of the hype.
## Build an evidence ledger, not a “safety score” [#build-an-evidence-ledger-not-a-safety-score]
Use a table you can audit later:
| Claim | Evidence | Checked at | Status | What would change it? |
| --------------------------- | --------------------------------------------- | ----------------- | ---------------------------- | ------------------------------------------- |
| Token can currently be sold | Two independent simulations on specified pool | Date/time + block | Pass / fail / unknown | Admin change, proxy upgrade, liquidity move |
| Main liquidity is locked | Locker contract and pool-position owner | Date/time + block | Verified / partial / unknown | Unlock, transfer, new active pool |
| Supply cannot expand | Verified implementation and all minter roles | Date/time + block | Verified / false / unknown | Role grant or upgrade |
| Team allocation vests | Vesting contract balances and schedule | Date/time + block | Verified / partial / unknown | Early release, migration, side wallets |
| Audit covers deployed code | Auditor report, commit, and address | Report date | Match / mismatch | Upgrade or redeployment |
This format prevents a common mistake: converting several partial checks into one permanent label.
## A hypothetical worked example [#a-hypothetical-worked-example]
Suppose Token Z has these results:
* Honeypot simulation succeeds with ordinary-looking tax today.
* Source is verified.
* The owner field is zeroed, but a separate role can pause transfers.
* 60% of the main v2 LP tokens are locked for 45 days.
* Two unlabeled wallets funded by the deployer hold 24% of supply.
* The audit covers an older implementation address.
A one-number scanner might still look favorable because the token is sellable and source is verified. CLEAR produces a different conclusion:
* **Contract:** unresolved pause authority.
* **Liquidity:** partial, short-duration lock.
* **Equity:** linked concentration.
* **Accountability:** audit mismatch.
* **Reality:** insufficient evidence until those contradictions are explained.
The correct output is not “scam proven.” It is **do not rely on the clean simulation; material risk remains unresolved**.
## What no rug pull checker can prove [#what-no-rug-pull-checker-can-prove]
* A team will not exploit a legitimate admin function later.
* Several wallets are economically independent.
* A private key or multisig signer will not be compromised.
* Locked liquidity will remain sufficient as price and volume change.
* The team will keep building.
* A partnership or roadmap claim is true.
* A market maker is not manipulating volume.
* A contract without known malicious patterns contains no novel trap.
* A currently sellable token will remain sellable after an upgrade.
Automated tools are strongest at repeatable technical checks. Human investigation remains necessary for identities, incentives, scope, contradictions, and changing control.
## How BlockMind fits—and where it does not [#how-blockmind-fitsand-where-it-does-not]
A BlockMind Pro agent can help organize a due-diligence investigation, inspect public sources, analyze holder and market context, retain an evidence trail in the [Notebook](/docs/memory/notebook), and monitor a tracked asset for material changes. A free [DeepDive report](https://blockmind.app/deepdive) can provide a structured first pass.
Neither is a safety certificate. Your agent can be wrong or outdated, does not replace a smart-contract audit, and cannot prove that an unidentified wallet belongs to a particular person. BlockMind never trades, moves funds, or tells you what to buy or sell.
For the wider workflow, pair this guide with the [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps), [honeypot token guide](/blog/honeypot-crypto-token), and [memecoin due-diligence checklist](/blog/memecoin-due-diligence-checklist).
## The Bottom Line [#the-bottom-line]
Use a crypto rug pull checker to find questions faster, not to end the investigation. Verify the address, run independent checks, inspect control and liquidity on-chain, map holder concentration, and test off-chain claims.
The most honest result is often “unknown” or “material issues unresolved.” That may feel less satisfying than a green badge, but it reflects what the evidence can actually support.
This article is for research and education, not financial advice. No checker or AI system can determine whether you should buy or sell a token.
## Sources [#sources]
1. [Chainalysis: Crypto market manipulation, wash trading, and pump-and-dump methodology](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/)
2. [Honeypot.is: Honeypot check API reference](https://docs.honeypot.is/ishoneypot)
3. [GoPlus: Token-security response fields](https://docs.gopluslabs.io/reference/response-details)
4. [Token Sniffer: Detection and API methodology](https://tokensniffer.readme.io/reference/introduction)
5. [Etherscan: What verified contract source means](https://info.etherscan.com/how-to-verify-contracts/)
6. [OpenZeppelin: Contract access control](https://docs.openzeppelin.com/contracts/5.x/access-control)
7. [Uniswap: Ownership of v2, v3, and v4 liquidity positions](https://support.uniswap.org/hc/en-us/articles/20980786685069-Why-is-liquidity-position-ownership-represented-by-tokens-or-NFTs)
8. [CFTC: Beware virtual-currency pump-and-dump schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
---
# How to Read the Crypto Fear & Greed Index (And When to Ignore It)
Published: 2026-02-07
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: crypto Fear and Greed Index
Keywords: crypto Fear and Greed Index, crypto sentiment, market psychology
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** The Fear & Greed Index is a useful gauge of market psychology, not a buy or sell signal. Extreme readings below 20 or above 80 have coincided with important turning points, but extremes can persist. Use it alongside [Bitcoin dominance](/blog/bitcoin-dominance-explained) and your own research, never as your sole decision-making tool.
***
The Fear & Greed Index measures crypto market sentiment on a scale of 0–100. Low scores mean fear and high scores mean greed. It's useful for understanding market psychology, but it won't tell you when to buy or sell.
## What the Numbers Mean [#what-the-numbers-mean]
| Score | Label | Market Mood |
| ------ | ------------- | ------------------------------------ |
| 0–24 | Extreme Fear | Panic, capitulation, possible bottom |
| 25–44 | Fear | Uncertainty, caution, risk-off |
| 45–55 | Neutral | Balanced, no strong bias |
| 56–74 | Greed | Optimism, confidence, risk-on |
| 75–100 | Extreme Greed | Euphoria, FOMO, possible top |
## How It's Calculated [#how-its-calculated]
Alternative.me combines several market and attention signals into its index:
| Active Input | What It Measures |
| ------------------- | ----------------------------------------------------- |
| Volatility | Price swings compared with recent historical averages |
| Momentum and volume | Trading volume and price trends |
| Social media | Crypto mentions, engagement, and interaction speed |
| Bitcoin dominance | Bitcoin's share of the crypto market |
| Google Trends | Changes in crypto-related search interest |
Alternative.me's [published methodology](https://alternative.me/crypto/fear-and-greed-index/) still lists surveys at 15%, but marks that input as "currently paused." Provider methods can change, so do not assume an old weight table describes today's calculation exactly. The durable idea is to aggregate several sentiment signals into one daily number.
## The Contrarian Approach [#the-contrarian-approach]
The classic wisdom: "Be fearful when others are greedy, and greedy when others are fearful."
In practice:
* **Extreme Fear (0–24):** A condition worth investigating for possible capitulation
* **Extreme Greed (75–100):** A condition worth reviewing for crowding and euphoria
This contrarian approach has historical support. Major market bottoms often coincide with extreme fear. Major tops often coincide with extreme greed.
### Historical Examples [#historical-examples]
**March 2020 (COVID crash):**
* Fear & Greed hit single digits
* Bitcoin dropped to $4,000
* Bitcoin's price rose roughly fifteenfold over the following 18 months
**November 2021 (market top):**
* Fear & Greed above 80 for weeks
* Bitcoin at $69,000
* Followed by 70%+ drawdown
## When to Ignore It [#when-to-ignore-it]
Here's where most guides stop. But the Fear & Greed Index has real limitations:
### 1. It Can Stay Extreme for Weeks [#1-it-can-stay-extreme-for-weeks]
Markets can remain "irrational" longer than you expect. Extreme fear doesn't mean the bottom is in. Extreme greed doesn't mean a crash is imminent.
**Example:** During the 2022 bear market, the index stayed in fear territory for months. Buying at "extreme fear" in May would have meant holding through another 50% drop.
### 2. It's a Lagging Indicator [#2-its-a-lagging-indicator]
The index measures current sentiment, not future price. By the time it shows extreme fear, prices have already fallen. By the time it shows extreme greed, prices have already risen.
### 3. Context Matters More Than the Number [#3-context-matters-more-than-the-number]
An extreme fear reading during a long-term bull market is different from extreme fear during a structural bear market. The number alone doesn't tell you which one you're in.
Combining Fear & Greed with [Bitcoin dominance](/blog/bitcoin-dominance-explained) gives you much better context. For instance, extreme greed plus falling BTC dominance can suggest late-cycle altcoin euphoria, historically a dangerous condition.
### 4. It Ignores Fundamentals [#4-it-ignores-fundamentals]
Sentiment can be wrong. The market might be euphoric for good reasons (genuine adoption, improving fundamentals). Or fearful for good reasons (protocol exploit, regulatory crackdown).
### 5. Short-Term Noise [#5-short-term-noise]
Daily readings fluctuate based on news cycles. A single tweet can spike or crash sentiment temporarily. Weekly or monthly trends are more meaningful than daily readings.
## How to Use It Properly [#how-to-use-it-properly]
### 1. Combine with Other Indicators [#1-combine-with-other-indicators]
Fear & Greed is one input, not the answer. Use it alongside:
* [BTC Dominance](/blog/bitcoin-dominance-explained) (risk appetite)
* On-chain metrics (actual holder behavior)
* Technical analysis (price levels)
* Your own research
### 2. Focus on Extremes [#2-focus-on-extremes]
The middle range (30–70) isn't very useful. Pay attention when readings are truly extreme, below 20 or above 80.
### 3. Look at Trends, Not Snapshots [#3-look-at-trends-not-snapshots]
A reading of 25 after weeks of 60+ is different from 25 after weeks of 15. The direction matters as much as the level.
### 4. Size Positions Accordingly [#4-size-positions-accordingly]
Rather than going all-in based on sentiment:
* Extreme fear: investigate whether the market is capitulating
* Extreme greed: review whether your exposure has become crowded
* Neutral: stick to your existing research process
### 5. Never Use It Alone [#5-never-use-it-alone]
The index tells you how others feel. It doesn't tell you if they're right. Do your own research.
## The Biggest Mistake [#the-biggest-mistake]
The biggest mistake is treating Fear & Greed as a trading signal. "The index is at 15, time to buy everything!"
Sentiment indicators help you understand the environment. They don't replace analysis, risk management, or patience.
**Use it to:**
* Gut-check your own emotions (Am I being greedy when others are?)
* Understand market context
* Identify possible turning points (not certainties)
**Don't use it to:**
* Time exact entries and exits
* Override fundamental analysis
* Make binary buy/sell decisions
During periods of extreme greed, weak projects can attract capital before investors examine the details. Read our guide on [5 signs of a rug pull](/blog/5-signs-crypto-rug-pull) before taking risk on an unfamiliar token.
## Check Market Sentiment with BlockMind [#check-market-sentiment-with-blockmind]
BlockMind refreshes the Fear & Greed Index once daily alongside the rest of its market context:
* **Current reading and history:** See today's value in the context of prior readings.
* **Related indicators:** Compare Fear & Greed with BTC Dominance and the Altcoin Season Index.
* **Contrarian Leverage Index:** Add BlockMind's in-house gauge built from derivatives positioning.
Create a free BlockMind account and open the [Market Indicators page](https://blockmind.app/market/indicators) to see Fear & Greed, BTC Dominance, Altcoin Season, and the Contrarian Leverage Index.
With a Pro trial or subscription, sentiment becomes context your personal AI investing agent watches for you. Its dashboard keeps the indicators in one carousel, and you can ask in chat how today's reading relates to your holdings. Your agent scans the market daily and brings material shifts into a [Morning Brief](/blog/crypto-morning-brief) delivered to your dashboard and email. It does research, not financial advice, and it will never tell you what to buy or sell.
Learn more about [how these indicators are calculated](/docs/classic/market-indicators).
## Frequently Asked Questions [#frequently-asked-questions]
### What is the Fear & Greed Index in crypto? [#what-is-the-fear--greed-index-in-crypto]
The Fear & Greed Index is a sentiment indicator that measures crypto market emotions on a scale of 0 (extreme fear) to 100 (extreme greed). Alternative.me currently describes volatility, momentum and volume, social media activity, Bitcoin dominance, and search trends as active inputs. It lists surveys as paused.
### Is a Fear & Greed score of 20 a good time to buy? [#is-a-fear--greed-score-of-20-a-good-time-to-buy]
Historically, extreme fear readings below 20 have often appeared near market bottoms. That relationship is not guaranteed because the market can stay fearful and drop further. Use it as one signal alongside other research, not a standalone buy trigger.
### How often is the Fear & Greed Index updated? [#how-often-is-the-fear--greed-index-updated]
Most providers update the index daily. However, daily readings can be noisy. Focus on weekly and monthly trends for more meaningful signals.
### Where can I track the Fear & Greed Index? [#where-can-i-track-the-fear--greed-index]
Create a free account and open BlockMind's Market Indicators page to see Fear & Greed alongside Bitcoin dominance, the Altcoin Season Index, and the Contrarian Leverage Index. All four refresh once daily.
## The Bottom Line [#the-bottom-line]
The Fear & Greed Index is one of the most accessible sentiment tools in crypto. Just remember:
* **It measures feelings, not facts**
* **Extremes matter more than mid-range readings**
* **Always combine it with other analysis**
* **It's a compass, not a GPS**
Get the full picture before making decisions, not just how the crowd feels today.
Place sentiment inside the [full crypto market analysis process](/blog/crypto-market-analysis-guide), and compare it with [on-chain activity](/blog/what-is-on-chain-analysis-crypto) before drawing a conclusion.
## Sources [#sources]
1. [Alternative.me: Crypto Fear & Greed Index methodology](https://alternative.me/crypto/fear-and-greed-index/), accessed July 2026.
2. [CFTC: Customer Advisory on AI Trading Bots and Investment Claims](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html), January 2024.
---
# What Is a Honeypot Crypto Token? How Sell Traps Work
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: informational
Primary keyword: honeypot crypto token
Keywords: honeypot crypto token, crypto honeypot checker, token cannot sell, honeypot token meaning, sell trap crypto
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A honeypot crypto token is designed or configured so people can buy but cannot sell normally—or can sell only after an abusive tax or restriction. A successful sell simulation is useful point-in-time evidence, not proof that the token will remain sellable. Never use real money merely to “test” a suspicious token.
***
The classic honeypot has an attractive chart and active buys. The trap appears only when a holder tries to exit. A contract can revert sales, restrict them to approved wallets, impose a near-total sell tax, or change its behavior after buyers arrive.
That is why “people are buying it” proves almost nothing. The contract can treat buys and sells differently, and visible transactions may come from privileged wallets, bots, or addresses controlled by the deployer.
## What does honeypot mean in crypto? [#what-does-honeypot-mean-in-crypto]
A honeypot token is a token whose contract, surrounding contracts, or configuration prevents ordinary holders from selling or extracts most of the value when they try.
The term usually covers several outcomes:
* a sell transaction always reverts;
* only allowlisted addresses can sell;
* blacklisted holders cannot transfer or sell;
* a maximum-sell rule makes a practical exit impossible;
* a dynamic sell tax captures nearly all proceeds;
* trading can be paused after purchases;
* a proxy upgrade adds restrictive behavior later;
* a router or pair-specific condition blocks the route ordinary users need.
[GoPlus' token-security documentation](https://docs.gopluslabs.io/reference/response-details) describes a honeypot signal as a token that may be unsellable because of contract functions or malicious code. It also allows an unknown result when no determination is available. That distinction matters: absence of a flag is not affirmative evidence of sellability.
## How a honeypot works [#how-a-honeypot-works]
On an automated market maker, a purchase and sale call the token contract through a router and liquidity pool. The token's transfer logic can inspect the sender, recipient, amount, block, or internal state and apply different rules.
### 1. Revert-on-sell logic [#1-revert-on-sell-logic]
The contract recognizes transfers to a known liquidity-pool address and rejects them. Wallet-to-wallet transfers may work, which can make casual checks look normal.
### 2. Allowlist or blacklist control [#2-allowlist-or-blacklist-control]
Selected addresses can trade while everyone else is blocked. The deployer can show successful sells from privileged addresses or let early trades work before changing the list.
### 3. Punitive taxes [#3-punitive-taxes]
The sale technically succeeds but a fee takes most or all of the output. [Honeypot.is](https://docs.honeypot.is/ishoneypot) documents simulated buy, sell, and transfer tax fields. A “successful” transaction with an extreme tax is economically equivalent to being trapped.
### 4. Maximum transaction constraints [#4-maximum-transaction-constraints]
A tiny maximum sell amount can force impractical numbers of transactions, each with gas and tax. Limits may be expressed in raw token units, making them harder to notice in a UI.
### 5. Switchable or upgradeable behavior [#5-switchable-or-upgradeable-behavior]
The token works during launch and becomes restrictive later. An owner, role, or proxy administrator may change tax, pause trading, replace the implementation, or modify exemptions.
OpenZeppelin's [access-control documentation](https://docs.openzeppelin.com/contracts/5.x/access-control) illustrates why checking only “owner renounced” is insufficient: systems can distribute authority across roles, multisigs, timelocks, or other contracts.
## The TRAP checklist [#the-trap-checklist]
Use **TRAP** to inspect the mechanism:
### T — Transfer path [#t--transfer-path]
* Which router and pool does the checker simulate?
* Does wallet-to-wallet transfer behave differently from a sale?
* Are there multiple pools with different rules?
* Does the token identify pool addresses dynamically or through a stored list?
### R — Roles and reconfiguration [#r--roles-and-reconfiguration]
* Who can change taxes, limits, exemptions, trading status, or pair addresses?
* Is the contract upgradeable?
* Which address controls the proxy implementation?
* Are privileged roles time-locked or controlled by a multisig?
### A — Access lists and amount limits [#a--access-lists-and-amount-limits]
* Are there blacklists, whitelists, cooldowns, anti-bot modes, or launch phases?
* What are maximum transaction and maximum wallet amounts?
* Can an admin add an ordinary holder after purchase?
* Are insiders or marketing wallets exempt from taxes and limits?
### P — Practical proceeds [#p--practical-proceeds]
* Does the sale simulation succeed?
* What amount returns after tax, slippage, and gas?
* Is liquidity deep enough for the position size?
* Does the result remain sensible for several trade sizes?
The practical question is not simply “did the call return?” It is “could an ordinary, non-privileged holder exit this realistic position through the active pool at a plausible cost?”
## How honeypot checkers work [#how-honeypot-checkers-work]
A simulator constructs a modeled buy and sell against a selected pair, executes the calls in a simulation environment, and reports whether they succeeded plus observed taxes, gas, and constraints.
Honeypot.is documents two important uncertainty states:
1. `simulationSuccess` can be false, in which case an error is returned.
2. `honeypotResult` or detailed simulation fields may be missing when the service cannot determine the result.
Some services also analyze holders or bytecode. Token Sniffer says its system compares source and bytecode with known scam patterns. These methods can catch traps without reproducing every route, but novel logic and changing state remain hard problems.
## The four-state result matrix [#the-four-state-result-matrix]
Do not flatten every output to green or red.
| Simulation | Contract/control review | Interpretation | Next step |
| ------------------------------------ | ------------------------------------------------------- | ------------------------------------------------------------- | ---------------------------------------------------------------------------- |
| Sell succeeds | No material reconfiguration found | Stronger point-in-time evidence, still not a safety guarantee | Verify liquidity, holders, audit scope, and team claims |
| Sell succeeds | Admin can change taxes, lists, pause, or implementation | Currently sellable but mutable | Identify controller, timelock, multisig, and change history |
| Sell fails | Clear restrictive logic or extreme tax | Material honeypot indicator | Do not use funds to retest; preserve evidence and stop |
| Simulation unknown/fails technically | Control unclear | No conclusion | Try an independent simulator and manual contract review; treat as unresolved |
This matrix is more honest than “score 87/100.” A sellable-but-mutable token and an immutable ordinary token should not receive the same interpretation just because both passed once.
## A worked hypothetical [#a-worked-hypothetical]
Assume a checker simulates buying 1 unit of the pool's base asset and reports:
* tokens received after buy: value equivalent to 0.96 units;
* modeled sale output before gas: 0.05 units;
* gas cost: 0.01 units;
* contract can change the sell tax;
* a deployer-linked wallet is tax-exempt.
Calculate the modeled recovery rate as `net sale output ÷ purchase input × 100`.
Here, net sale output is `0.05 - 0.01 = 0.04`. The recovery rate is `0.04 ÷ 1 × 100 = 4%` before considering price movement. The transaction “works,” but the position is economically trapped. The mutable tax and exempt insider wallet are additional risks.
This hypothetical is an explanation, not a universal threshold. Ordinary taxes, slippage, and gas vary by chain, pool, and trade size.
## A safe verification workflow [#a-safe-verification-workflow]
### Step 1: establish the exact address [#step-1-establish-the-exact-address]
Copy the contract or mint address from official project documentation and cross-check it. Names and symbols can be duplicated.
### Step 2: run two independent simulations [#step-2-run-two-independent-simulations]
Use tools with documented methodologies, such as Honeypot.is and a GoPlus-powered interface. Match chain and pair. Save timestamps and outputs.
### Step 3: inspect verified source and implementation [#step-3-inspect-verified-source-and-implementation]
On an explorer, confirm source verification, proxy status, implementation address, creator, privileged roles, and recent administrative calls. Etherscan says verification matches source code to deployed bytecode, but it does not certify the code as secure.
### Step 4: read the result in economic terms [#step-4-read-the-result-in-economic-terms]
Compare buy and sell tax, gas, maximum amounts, pool depth, and slippage for realistic sizes. A sale worth almost nothing is not a meaningful exit.
### Step 5: check mutability [#step-5-check-mutability]
Find who can change the rules. Review role grants, ownership transfers, proxy upgrades, tax changes, and pause events.
### Step 6: continue broader due diligence [#step-6-continue-broader-due-diligence]
A non-honeypot token can still rug through liquidity removal, insider dumping, minting, false claims, or abandonment. Continue with the [crypto rug pull checker workflow](/blog/crypto-rug-pull-checker-guide), [liquidity locks guide](/blog/liquidity-locks-crypto-explained), and [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps).
## What not to do [#what-not-to-do]
* Do not buy a small amount just to see whether you can sell it.
* Do not connect a valuable wallet to an unfamiliar checker.
* Do not sign token approvals or arbitrary messages to run a public scan.
* Do not trust a screenshot supplied by the project; reproduce the check.
* Do not assume visible sells came from ordinary wallets.
* Do not interpret “not a honeypot” as “not a scam.”
* Do not rely on a ticker when entering the address.
A public checker generally needs a contract address, not your seed phrase, private key, or a wallet signature.
## Can a honeypot be bypassed? [#can-a-honeypot-be-bypassed]
An ordinary buyer should not assume so. Some traps depend on routes or amounts, but trying to outmaneuver malicious code can expose more funds, approvals, or gas while alerting the operator. A scam contract can also change state between attempts.
If funds are already affected, preserve transaction hashes, addresses, URLs, and communications. Revoke unnecessary token approvals through a reputable wallet or explorer interface, contact the exchange or wallet provider involved, and report suspected fraud to the appropriate authorities. Avoid “recovery” strangers asking for an upfront fee or seed phrase.
## Limitations and counterevidence [#limitations-and-counterevidence]
A failed simulation does not always prove malicious intent. Liquidity may be absent, a router unsupported, a pool paused, or a node call may fail. That is why the correct initial label can be “unknown.” Conversely, a successful simulation can become stale immediately after a role change or proxy upgrade.
Legitimate projects sometimes use transfer limits, pause functions, allowlists, or anti-bot rules during launch. Those controls still create centralization and exit risk. The relevant questions are who controls them, why they exist, when they end, and whether the public explanation matches the deployed code.
## How BlockMind can support the research [#how-blockmind-can-support-the-research]
A BlockMind agent can help you compare scanner outputs, inspect public documentation, save contract addresses and claims in the [Notebook](/docs/memory/notebook), and monitor a tracked asset for changes. It cannot certify a token, replace a contract auditor, or know hidden wallet ownership. It never trades or moves funds.
## The Bottom Line [#the-bottom-line]
A honeypot token turns apparent market access into a one-way door. Use simulators to test the door without funds, then inspect who can change the lock.
Treat pass, fail, and unknown as distinct results. A pass is point-in-time evidence. A fail is a serious warning. Unknown means keep your capital out of the experiment until the uncertainty is resolved.
This article is for research and education, not financial advice.
## Sources [#sources]
1. [Honeypot.is: Honeypot check and simulation response](https://docs.honeypot.is/ishoneypot)
2. [GoPlus: Token-security response details](https://docs.gopluslabs.io/reference/response-details)
3. [Token Sniffer: Contract and bytecode detection methodology](https://tokensniffer.readme.io/reference/introduction)
4. [Etherscan: Verifying contract source code](https://info.etherscan.com/how-to-verify-contracts/)
5. [Ethereum.org: Smart-contract verification](https://ethereum.org/developers/docs/smart-contracts/verifying/)
6. [OpenZeppelin: Access control and administrative roles](https://docs.openzeppelin.com/contracts/5.x/access-control)
---
# How AI Accesses Real-Time Crypto Market Data
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: informational
Primary keyword: how AI accesses real-time crypto market data
Keywords: how AI accesses real-time crypto market data, real-time crypto AI, crypto market data API AI, live crypto data sources, AI crypto data pipeline
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** AI does not inherently know a live crypto price. A useful system calls external sources—exchange WebSockets, market-data APIs, blockchain nodes and indexers, news or social sources, and read-only portfolio connections—then normalizes and timestamps the results before a model explains them. “Real time” is not one standard: a trade feed may update event by event, an aggregator may cache for seconds, a blockchain value may wait for confirmation, and a daily indicator may update once per day.
The language model is the interpretation layer, not the market-data feed. If its tools are disabled, a current-looking answer may come from training memory. If search is enabled, it may find a recent webpage. If structured connections are available, it can query a defined endpoint and receive machine-readable data. Those three paths have very different freshness and reliability.
This data-provenance view supports the broader division of labor in [AI vs manual crypto research](/blog/ai-vs-manual-crypto-research): let systems retrieve and normalize evidence, while people inspect assumptions and decide what the evidence means.
## The six layers between a market event and an AI answer [#the-six-layers-between-a-market-event-and-an-ai-answer]
### 1. A venue or blockchain produces an event [#1-a-venue-or-blockchain-produces-an-event]
For centralized markets, an exchange matches an order and emits a trade or order-book update. For a blockchain, a transaction enters a block and changes state. For news, an issuer, regulator, developer, or reporter publishes information.
There is no single universal “crypto price.” BTC/USD on one venue can differ from BTC/USDT on another. A decentralized-exchange pool can show another executable price, especially for a large order. Before analysis begins, the system must know the asset, pair, venue, currency, and event type.
### 2. A feed transports the data [#2-a-feed-transports-the-data]
Streaming systems commonly use WebSockets for frequent updates. Coinbase’s official Exchange documentation describes a public WebSocket feed for orders and trades, but it also warns that messages can be dropped or arrive out of order; clients use sequence numbers to detect gaps ([Coinbase](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)).
Request-response APIs are better for snapshots and slower-changing datasets. CoinGecko’s `/simple/price` endpoint can return price, market cap, 24-hour volume, change, and a `last_updated_at` timestamp. Its documentation states a 20-second cache/update frequency for listed Pro API plans ([CoinGecko](https://docs.coingecko.com/reference/simple-price)). That is fresh enough for many research questions, but it is not tick-by-tick execution data.
On Ethereum, applications read chain state through a node’s JSON-RPC interface. Queries can specify `latest`, `safe`, `finalized`, or an exact block number, which makes “current” a choice with different finality properties ([Ethereum.org](https://ethereum.org/developers/docs/apis/json-rpc/)).
### 3. An indexer turns raw events into useful tables [#3-an-indexer-turns-raw-events-into-useful-tables]
Raw chain data is difficult to query at scale. Indexers decode contract events, organize addresses and transactions, and calculate derived metrics. Ethereum’s analytics documentation describes core raw tables for blocks, transactions, logs, and traces, with higher-level protocol tables built after decoding ([Ethereum.org](https://ethereum.org/developers/docs/data-and-analytics/)).
This layer is where useful concepts appear:
* token transfers and balances;
* decentralized-exchange swaps;
* protocol deposits and withdrawals;
* fees and revenue estimates;
* bridge flows;
* holder rankings;
* labeled exchange, fund, protocol, or “smart money” addresses.
Derived does not mean false, but it does mean methodological. “Active user,” “protocol revenue,” and “whale” depend on definitions. A careful AI answer names the provider or method rather than presenting every metric as an objective property of the chain.
The distinction between a raw record, a decoded action, and an entity-level inference is developed in [what on-chain analysis means](/blog/what-is-on-chain-analysis-crypto).
### 4. A normalization layer reconciles symbols, time, and units [#4-a-normalization-layer-reconciles-symbols-time-and-units]
This quiet step prevents many confident errors. A system needs to:
* map tickers to canonical asset and contract identifiers;
* distinguish native, wrapped, and bridged assets;
* convert quote currencies without mixing timestamps;
* normalize token decimals;
* avoid summing the same holding across duplicate sources;
* align hourly, daily, and block-level series;
* preserve venue and chain provenance;
* apply a consistent timezone.
An AI model should not improvise these joins from prose. Reliable systems pass it a structured result with identifiers and metadata.
### 5. Retrieval selects evidence for the question [#5-retrieval-selects-evidence-for-the-question]
The system routes each question to the right source. “What is the last BTC trade on Coinbase?” needs a venue feed. “How concentrated is this ERC-20?” needs the correct contract, holder balances, and address classification. “Why did it move?” may need news, market, derivatives, on-chain, and social evidence.
A general AI with web search can retrieve current pages and citations. OpenAI’s current product documentation confirms that ChatGPT can search the web for current information ([OpenAI](https://help.openai.com/en/articles/12677804-what-is-chatgpt-faq)). Search is valuable, but a web page is not equivalent to a guaranteed price feed. The page may cache data, omit a timestamp, quote another venue, or refer to a similarly named asset.
### 6. The model explains the result [#6-the-model-explains-the-result]
Only now does generative AI add value: it compares sources, translates technical data, relates a change to a portfolio or thesis, and states uncertainty. The answer should retain the metadata instead of stripping it away.
For a time-sensitive number, expect at least:
> BTC/USD was X on venue Y at timestamp Z. This is a last-trade/mark/index/aggregate price from source S. Data was retrieved at T.
If a product only returns “Bitcoin is X,” you do not know enough to evaluate freshness.
Apply this quote checklist to a common interface in [Can ChatGPT give accurate crypto prices?](/blog/chatgpt-crypto-prices-accuracy).
## “Real time” has five timestamps [#real-time-has-five-timestamps]
The phrase becomes clearer when you track the full chain:
1. **Event time:** when the trade, block, post, or filing occurred.
2. **Source time:** when the provider recorded or published it.
3. **Ingestion time:** when the application received it.
4. **Analysis time:** when calculations or the AI response ran.
5. **Display time:** when you saw the answer.
A system can retrieve data instantly and still show an old event. It can also analyze a fresh event but display it after a long-running report. “Updated just now” usually describes one of these timestamps, not all five.
## Match freshness to the decision [#match-freshness-to-the-decision]
| Use case | Appropriate source | Typical freshness concept | Main failure mode |
| ---------------------- | ----------------------------------------- | -------------------------------- | --------------------------------------------- |
| Explain a token | Official docs and current web sources | Publication date | Outdated or marketing-led documentation |
| Portfolio overview | Read-only balances plus normalized prices | Last balance and price refresh | Missing venue, derivative, or duplicate asset |
| Price alert | Venue or aggregate price feed | Tick or seconds | Wrong pair, transient wick, stale connection |
| Order-book analysis | Exchange WebSocket and snapshot recovery | Sequence-aware stream | Dropped messages and phantom depth |
| Holder distribution | Chain indexer at stated block | Block height/finality | Misclassified exchanges and contracts |
| Protocol fundamentals | Decoded chain data | Provider methodology and refresh | Inconsistent definitions or double counting |
| Market sentiment | Social/news data | Window and collection time | Bots, sampling changes, price-following tone |
| Daily market indicator | Scheduled calculated series | Daily observation | Calling a daily signal “live” |
BlockMind’s classic market indicators, for example, update once daily and the documentation says they should be treated as daily market-temperature readings—not live feeds ([market indicator documentation](/docs/classic/market-indicators)). That is not a defect; it is the correct cadence for that product. The problem would be describing the same figure as tick-level.
## How portfolio-aware AI gets current holdings [#how-portfolio-aware-ai-gets-current-holdings]
A portfolio-aware system needs permissioned, read-only inputs:
* public wallet addresses;
* balance-only exchange connections;
* imported structured files; or
* supported portfolio integrations.
It then resolves balances, prices assets, groups exposures, and attaches the result to relevant questions. It should never require a seed phrase, private key, withdrawal permission, or transaction-signing authority for research. BlockMind’s connection guide documents the current supported mechanics and scope in one place: [connect your portfolio](/docs/getting-started/connect-your-portfolio).
General-purpose AI can also analyze an uploaded CSV or spreadsheet. OpenAI notes, however, that its data-analysis environment cannot make external web requests; external data must be uploaded or connected before analysis ([OpenAI](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt/)). A static export therefore goes stale until you replace it.
## A freshness checklist for any crypto AI answer [#a-freshness-checklist-for-any-crypto-ai-answer]
Ask these questions before relying on a current claim:
1. What exact asset, chain, and contract address was resolved?
2. Which venue, pair, or aggregation method produced the price?
3. Is the value a last trade, midpoint, mark, index, close, or estimated execution price?
4. What are the event and retrieval timestamps?
5. Is the endpoint streamed, polled, cached, or scheduled?
6. Were sequence gaps, rate limits, or partial failures detected?
7. What block height or finality tag was used for chain data?
8. Which figures are raw and which are provider-derived?
9. Were conflicting sources reconciled or merely averaged?
10. What data was unavailable?
## Limitations and counterevidence [#limitations-and-counterevidence]
More feeds do not automatically create a better answer. They can create contradictory timestamps, duplicate assets, and false precision. Low-latency data also does not guarantee executable liquidity: visible quotes can move before an order fills.
Other limits remain:
* Exchange and API outages can interrupt fresh data.
* Rate limits and caches create intentional delays.
* Blockchain data may be reorganized before finality.
* Centralized-exchange customer trades and balances are not fully visible on-chain.
* Address labels are estimates and can become outdated.
* Web sources can change after an answer is generated.
* A model can misread perfectly fresh structured data.
The right design preserves source provenance, detects missing data, and says “unknown” rather than filling a gap with plausible text.
## The Bottom Line [#the-bottom-line]
AI accesses current crypto data through external tools. The trustworthy unit is not “an AI answer”; it is a traceable chain from event to feed, normalization, retrieval, and explanation. Demand an asset identifier, source, venue or chain, timestamp, metric definition, and stated limitations.
BlockMind combines market research, on-chain sources, web research, and read-only portfolio context in your agent’s workflow. It is still research, not financial advice: the agent never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [Coinbase Developer Documentation — Exchange WebSocket Overview](https://docs.cdp.coinbase.com/exchange/websocket-feed/overview)
* [CoinGecko API — Simple Price Endpoint](https://docs.coingecko.com/reference/simple-price)
* [Ethereum.org — JSON-RPC API](https://ethereum.org/developers/docs/apis/json-rpc/)
* [Ethereum.org — Data and Analytics](https://ethereum.org/developers/docs/data-and-analytics/)
* [OpenAI — What Is ChatGPT?](https://help.openai.com/en/articles/12677804-what-is-chatgpt-faq)
* [OpenAI — Data Analysis with ChatGPT](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt/)
---
# How Long Does It Take to Research a Crypto Token? Manual vs AI Agent
Published: 2026-03-10
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: informational
Primary keyword: how long does it take to research a crypto token
Keywords: crypto research time, research a crypto token, crypto due diligence
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** BlockMind's illustrative baseline for a disciplined manual first pass is **about 2–3 hours per token**: 30 minutes on the team, 20 on tokenomics, 45 on on-chain analysis, 30 on social sentiment, and 20 on technicals. This is a planning estimate, not a measured universal average; the token, available evidence, and depth of review can move it sharply.
***
If you're asking how long to research crypto properly, the honest answer is: longer than most people think. Scanning a chart for five minutes is not research. Reading a few posts on X is not research. Real due diligence means checking whether the team is credible, whether the token economics make sense, whether wallets are accumulating or dumping, whether the social buzz is real, and whether the chart is confirming or contradicting the story.
That time cost matters because crypto remains full of asymmetric risk. In its 2026 Crypto Crime Report, Chainalysis estimated that scams and fraud received at least $14 billion on-chain in 2025 and projected that the figure could exceed $17 billion as more addresses are identified. In a market like that, rushed research is expensive research. Source: [Chainalysis 2026 Crypto Crime Report](https://www.chainalysis.com/blog/crypto-scams-2026/).
## An illustrative baseline for manual crypto research [#an-illustrative-baseline-for-manual-crypto-research]
For this guide, we use **2–3 hours** as an illustrative baseline for a disciplined first pass by hand. It is not an empirical industry average. A familiar, well-documented asset may take less; a new, obscure, or technically complex project can take much longer.
That estimate assumes a fairly disciplined process:
* **Team check:** 30 minutes
* **Tokenomics review:** 20 minutes
* **On-chain analysis:** 45 minutes
* **Social sentiment review:** 30 minutes
* **Technical analysis:** 20 minutes
That already adds up to **145 minutes**, or **2 hours and 25 minutes**, before you write a conclusion, compare alternatives, or double-check anything confusing.
For many investors, the real bottleneck is not willingness. It is fragmentation. Manual DYOR usually means bouncing between a block explorer, token unlock dashboards, social feeds, charting tools, exchange listings, and the project's own site. You are not just analyzing the token. You are stitching together context from six different places and hoping you miss nothing important.
## Why does a team check alone take around 30 minutes? [#why-does-a-team-check-alone-take-around-30-minutes]
In this baseline, a team check gets **about 30 minutes** because credibility is rarely obvious from the homepage.
You need to verify whether real people are behind the project, whether they have relevant experience, and whether their claims hold up outside their own marketing. That often means checking LinkedIn profiles, prior startups, GitHub activity, conference appearances, interviews, and whether the same names show up consistently across the website, whitepaper, and community channels.
The hard part is not finding flattering information. It is finding disconfirming information. If a founder claims to have built successful protocols before, can you verify that? If the team is anonymous, is there a good reason, or is anonymity being used as a shield? If there is no history, no public reputation, and no accountability, that changes the risk profile immediately.
If you want a faster checklist for obvious danger signals, read [5 signs a crypto project might be a rug pull](/blog/5-signs-crypto-rug-pull).
## How much time should tokenomics research take? [#how-much-time-should-tokenomics-research-take]
This baseline gives tokenomics **around 20 minutes for a first-pass review**, assuming the project actually discloses the key numbers clearly.
In that window, you are trying to answer a few basic questions fast:
* What is the current circulating supply versus max supply?
* How much is allocated to team, investors, treasury, and community?
* When are major unlocks scheduled?
* Does the token have real utility, or is it mostly narrative packaging?
* Could new supply hit the market hard over the next few months?
This is where many retail investors underestimate the work. A token can look attractive on price alone while the supply structure is quietly working against you. A project with aggressive insider allocations, near-term vesting cliffs, or vague treasury rules can create sell pressure even if the story sounds strong.
Useful manual sources here often include the token's official docs, vesting trackers like [Tokenomist](https://tokenomist.ai/), and market data sites that separate circulating and fully diluted valuation.
## Why does on-chain analysis take the longest? [#why-does-on-chain-analysis-take-the-longest]
This baseline gives on-chain analysis **45 minutes or more** because it is the part most likely to reveal what the market is actually doing, not just what the project says.
A serious on-chain pass can include:
* Checking top holder concentration
* Looking for whale accumulation or distribution
* Reviewing transfer activity around unlocks or major news
* Seeing whether liquidity looks healthy
* Comparing wallet behavior with the public narrative
This is also where manual research gets slow fast. Even if you know how to use explorers, wallet labels, and dashboard tools, there is a difference between seeing transactions and interpreting them. Ten large inflows might be bullish, or they might be exchange reshuffling. A concentrated holder base might be normal for an early-stage protocol, or it might mean one exit can crush the market.
The reason this step matters is simple: on-chain activity is one of the few sources in crypto that lets you observe behavior directly. If you skip it, you are relying heavily on narratives. A structured first-pass report can organize the evidence, but it does not replace verification at the explorer and project sources.
## How long does social sentiment analysis take? [#how-long-does-social-sentiment-analysis-take]
This baseline gives social sentiment analysis **about 30 minutes** if you want signal instead of noise.
That time is not about counting followers. It is about figuring out whether attention is organic, informed, and durable. You are looking for things like:
* Is the discussion driven by users or by incentive campaigns?
* Are respected analysts discussing the project, or only promoters?
* Is engagement increasing because of product traction, or because price just moved?
* Does community sentiment match what on-chain and tokenomics data suggest?
This step is easy to do badly. Social feeds can make weak projects look inevitable. High engagement can reflect speculation, bots, or short-lived meme energy rather than durable conviction.
Sentiment is most useful when paired with broader market context. For example, when the market is euphoric, weak tokens can look stronger than they are. Our guide on [how to read the Crypto Fear & Greed Index](/blog/fear-greed-index-guide) explains why sentiment should be treated as context, not proof.
## How much time should technical analysis take? [#how-much-time-should-technical-analysis-take]
This baseline gives technical analysis **around 20 minutes for a practical first pass**, not because charts are unimportant, but because they are usually the final layer, not the first one.
A sensible technical review is often enough to answer:
* Where are the obvious support and resistance levels?
* Is momentum confirming the narrative?
* Did the token just make an unsustainable breakout?
* Is volume supporting the move?
* Are you chasing after a large expansion candle?
Charts help with timing. They do not replace project research. A clean chart cannot rescue broken tokenomics, and strong momentum does not make a weak team trustworthy.
For most investors, the right role of technicals is simple: use them to improve entries, exits, and risk management **after** you understand the asset.
## Why does manual DYOR feel so slow in practice? [#why-does-manual-dyor-feel-so-slow-in-practice]
Manual DYOR feels slow because each research layer depends on a different tool, a different format, and a different type of judgment.
You are switching between founders, unlock schedules, wallet behavior, sentiment feeds, and charts. Then you still need to synthesize all of that into one decision: *Is this token strong, risky, overpriced, early, late, or just noisy?*
That is exactly why most people cut corners. They skip the team check. They ignore unlocks. They treat social hype as validation. They look at price action and call it conviction.
The result is not just bad accuracy. It is inconsistency. One week you spend three hours on a token. The next week you spend twelve minutes and convince yourself it was enough.
This is exactly the pain a personal AI analyst is built to solve. If the full manual process takes hours per token, the useful question is which parts you should hand to an agent that gathers and structures the data, and which parts still need your judgment. [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research) examines that split in detail.
## What should you do if you cannot spend 2–3 hours on every token? [#what-should-you-do-if-you-cannot-spend-23-hours-on-every-token]
You should split the work: **hand the data-gathering hours to an agent, and keep the judgment calls for yourself**.
Look back at the 145-minute breakdown. Most of it involves pulling holder data, checking unlocks, scanning sentiment, and assembling market structure. That is gathering and stitching, not judging, and it is exactly the work a machine should do.
This is what BlockMind is built around. On Pro, BlockMind gives you a **personal AI investing agent** with its own workspace that carries the gathering while you carry the decision. Instead of six tabs, the work follows four clear steps in the [research journey](/docs/capabilities/research-journey):
1. **Explore** scans the market by asset class, timeframe, and market cap, then turns the results into saved ideas for a fast first-pass screen.
2. **Analyze** takes an idea that survived the screen and produces a structured analysis: conviction, levels, competitors, valuation. The 45-minute stitching pass becomes one request.
3. **Verdicts** gets second opinions from a panel of experts covering macro conditions, charts, risk, and trade readiness. The result is a clear call: approved, watch, wait, or reject.
4. **Track** keeps watching after you decide. Your agent can [monitor your positions and tracked assets](/docs/automations/monitoring-and-alerts), and every morning a [Morning Brief](/docs/automations/morning-brief) on your dashboard and in your email inbox explains what moved and why it matters to your holdings.
Because everything is grounded in your [connected portfolio](/docs/getting-started/connect-your-portfolio) and remembered in a shared Notebook, next month's research builds on this month's instead of starting from zero.
One thing does not change: your agent does research, not financial advice. It will never tell you what to buy or sell. BlockMind uses wallet and exchange connections only to read balances and positions, and the agent cannot trade, withdraw, or move funds even if you told it to. It compresses the 145 minutes of gathering; the judgment stays yours, which is exactly where it belongs.
The agent comes with Pro; see [current plans and introductory terms](/docs/overview/plans-and-pricing). If you just want to try structured token research first, [DeepDive reports](https://blockmind.app/deepdive) are free, up to five per day per email address.
## Frequently Asked Questions [#frequently-asked-questions]
### How long does it take to research crypto? [#how-long-does-it-take-to-research-crypto]
BlockMind's illustrative baseline allows **about 2–3 hours per token** for a disciplined manual first pass across the team, tokenomics, on-chain activity, social sentiment, and technicals. It is a planning estimate rather than a universal measured average.
### How much time does one token take to research? [#how-much-time-does-one-token-take-to-research]
In BlockMind's illustrative checklist, the allocations add up to **145 minutes**, or **2 hours and 25 minutes**, for a disciplined first-pass manual review. It is not a universal benchmark; complex or early-stage tokens can take longer, while familiar assets may take less.
### Can I research a crypto token in 10 minutes? [#can-i-research-a-crypto-token-in-10-minutes]
You can do a quick screen in 10 minutes, but you cannot do proper due diligence in 10 minutes. That is enough time to catch obvious red flags, not to understand the full risk and opportunity profile.
### What takes the most time in crypto token research? [#what-takes-the-most-time-in-crypto-token-research]
On-chain analysis usually takes the most time because it requires checking wallet behavior, holder concentration, liquidity, and transaction patterns, then interpreting what those signals actually mean.
### Does BlockMind replace manual crypto research? [#does-blockmind-replace-manual-crypto-research]
BlockMind does not remove judgment. Your personal AI investing agent researches a token across the layers in this post, writes an analysis report, and can put it in front of a panel of experts for a verdict. You still make the call; you just start from a finished brief instead of twelve open tabs.
## The Bottom Line [#the-bottom-line]
If you want the honest answer to how long it takes to research a crypto token properly, it is this: **long enough that most people do not do it consistently by hand**.
That is the real problem. The issue is not that investors do not care. It is that good research is fragmented, repetitive, and time-consuming. When a careful first pass can take hours per token, an agent doing the gathering stops being a luxury and starts becoming the difference between disciplined investing and expensive guesswork.
## Keep reading [#keep-reading]
- [AI vs Manual Crypto Research](https://docs.blockmind.app/blog/ai-vs-manual-crypto-research)
- [The research journey](https://docs.blockmind.app/docs/capabilities/research-journey)
- [DeepDive reports](https://docs.blockmind.app/docs/classic/deepdive-reports)
## Sources [#sources]
1. [Chainalysis: 2026 Crypto Scam Trends](https://www.chainalysis.com/blog/crypto-scams-2026/), January 2026.
2. [Tokenomist: Token Unlocks and Vesting Data](https://tokenomist.ai/), accessed July 2026.
3. [SEC Investor.gov: Crypto Asset Scam Warning Signs](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/crypto-scams), May 2024.
---
# How to Check If a Crypto Team Is Legitimate: Evidence Checklist
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: how to check if a crypto team is legitimate
Keywords: how to check if a crypto team is legitimate, verify crypto team, crypto founder background check, crypto project team red flags, crypto due diligence team
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A legitimate-looking team page is not verification. Check four separate claims: the people are who they say they are, they have evidence of relevant work, a real entity or governance structure is accountable, and the people controlling code, upgrades, treasury, liquidity, and communications match the public story. Build independent evidence links; never treat LinkedIn, a conference photo, an audit logo, or a “Verified” social badge as proof by itself.
Known founders can still fail or commit misconduct, and anonymous developers can build legitimate open-source systems. Team verification does not predict success. Its job is narrower: detect fabrication, unexplained conflicts, capability gaps, and hidden control before they become your risk.
Team evidence is one part of the complete [what to check before buying crypto](/blog/what-to-check-before-buying-crypto) process, alongside tokenomics, ownership, liquidity, security, and usage.
## The four questions behind team legitimacy [#the-four-questions-behind-team-legitimacy]
### 1. Identity: Do these people exist, and are the profiles connected to them? [#1-identity-do-these-people-exist-and-are-the-profiles-connected-to-them]
### 2. Capability: Have they done work relevant to the claims? [#2-capability-have-they-done-work-relevant-to-the-claims]
### 3. Accountability: What legal, governance, or reputational consequences apply? [#3-accountability-what-legal-governance-or-reputational-consequences-apply]
### 4. Control: Who can actually change the software or move the assets? [#4-control-who-can-actually-change-the-software-or-move-the-assets]
A project can pass one and fail another. A real founder with a strong résumé does not make an upgrade key safe. An anonymous team may have a long, verifiable code history and constrained governance. Keep the dimensions separate.
## Step 1: Freeze the team’s claims [#step-1-freeze-the-teams-claims]
Save dated copies or links to:
* official team and foundation pages;
* whitepaper and disclosures;
* legal entity names and jurisdictions;
* founder and executive biographies;
* repository organizations;
* audit and security pages;
* governance documentation;
* treasury, multisig, and admin addresses;
* official social accounts and contact domains.
Record exact claims: employer, dates, role, credential, product shipped, amount raised, partner, investor, audit, and governing entity. Vague claims are hard to disprove by design.
## Step 2: Verify identity through independent paths [#step-2-verify-identity-through-independent-paths]
For each public team member, seek evidence that predates the project:
* profile history with consistent dates;
* old repository commits, talks, articles, patents, or publications;
* former employer pages or archived announcements;
* interviews hosted by independent organizations;
* prior projects with users, releases, or verifiable outcomes;
* public interactions with established peers over time.
Cross-link the evidence. A project site linking to a profile and that profile linking back to the project is one self-asserted loop, not two independent confirmations.
### Reverse-image and media checks [#reverse-image-and-media-checks]
Search headshots and key promotional frames for unrelated identities or stock images. Examine whether videos originate from the claimed account and event. Deepfakes and stolen media make visual familiarity weak evidence.
The CFTC specifically recommends researching personnel, running reverse-image searches, and checking domain registration history before trusting AI or crypto trading promotions ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)).
### Watch for résumé inconsistencies [#watch-for-résumé-inconsistencies]
* overlapping full-time roles with no explanation;
* titles that grow more senior across newer biographies;
* a former employer that never mentions the person;
* credentials with no institution, year, or record;
* one biography copied across supposed team members;
* profile creation or activity beginning immediately before fundraising;
* followers and endorsements disconnected from the claimed field.
Absence is not proof of fraud. Record it as unverified and adjust confidence.
## Step 3: Verify the organization and domain [#step-3-verify-the-organization-and-domain]
Find the legal entity behind the website, token sale, terms, employment, trademarks, and treasury. Depending on jurisdiction, check official corporate registries and regulator databases. Ask:
* Does the entity exist and remain active?
* When was it formed relative to fundraising?
* Do directors or beneficial owners match disclosures where public?
* Which entity owes users obligations?
* Which law and dispute forum do the terms name?
* Are token issuer, foundation, developer company, and marketing entity different?
Check the domain through ICANN’s RDAP lookup. ICANN explains that RDAP provides current registration data and replaced WHOIS as the standard protocol, while privacy rules may redact personal fields ([ICANN](https://www.icann.org/en/contracted-parties/registry-operators/registration-data-access-protocol/information-for-rdap-users-31-08-2018-en)).
Domain age is a clue, not a verdict. A new domain can belong to a real new project; an old domain can be acquired by scammers. Look for registrant or registrar changes, look-alike spelling, inconsistent email domains, certificate history, and archived content.
## Step 4: Verify capability in the repository [#step-4-verify-capability-in-the-repository]
Do not count commits and stop. Inspect:
* who authored substantive code;
* whether core developers match the public team;
* release tags and deploy history;
* dependency and fork history;
* issue and pull-request discussion quality;
* whether development is ongoing or imported in one dump;
* test, security, and review practices;
* the relationship between public code and deployed contracts.
GitHub’s contributor graph can show top contributors to the default branch, but GitHub notes it has scope limits and does not count some contributions, such as unmerged work or certain commits ([GitHub](https://docs.github.com/en/repositories/viewing-activity-and-data-for-your-repository/viewing-a-projects-contributors)). A sparse public profile may hide private or unmerged work; a dense graph may contain trivial changes.
Signed commits provide evidence about a cryptographic key associated with an account. GitHub’s “Verified” status means the signature was successfully verified; it does not certify code quality, employer history, or that every claimed author personally wrote the change ([GitHub](https://docs.github.com/en/authentication/managing-commit-signature-verification/about-commit-signature-verification)).
## Step 5: Verify shipped work, not roadmap language [#step-5-verify-shipped-work-not-roadmap-language]
For each major claim, look for a live artifact:
* deployed contract at a documented address;
* product users can access;
* release matching repository code;
* protocol activity consistent with the claimed use;
* governance proposals and execution;
* independent integrations confirmed by both parties;
* incident history and postmortems.
A slick demo can be staged. A contract can exist with no meaningful users. A high transaction count can be incentives or bots. Combine artifacts rather than elevating one.
For “partnership” claims, seek an independent announcement from the named partner and clarify whether it means a paid vendor relationship, technical integration, investment, pilot, or marketing campaign.
## Step 6: Trace who controls the system [#step-6-trace-who-controls-the-system]
This is often more important than biographies. Identify:
* contract owner and upgrade admin;
* proxy implementation and change process;
* pause, freeze, mint, blacklist, fee, and rescue powers;
* treasury and liquidity multisigs;
* signer identities and threshold;
* timelocks and emergency bypasses;
* governance concentration and delegation;
* front-end domain and deployment control;
* social, repository, package-registry, and DNS administrators.
Compare public decentralization claims with actual control. A nine-member advisory board does not matter if one person controls a 1-of-1 upgrade key and treasury.
Use the holder workflow in [How to check token holder distribution](/blog/crypto-holder-distribution) to separate team, vesting, treasury, exchange, and infrastructure balances.
## Step 7: Read audits in scope [#step-7-read-audits-in-scope]
An audit logo does not mean “safe.” Open the report and verify:
* auditor and report URL;
* repository commit or source files reviewed;
* contract addresses, if deployment was reviewed;
* audit date relative to later changes;
* severity definitions;
* unresolved and acknowledged findings;
* whether fixes were retested;
* excluded components and assumptions;
* admin-key and economic-design observations.
Confirm the auditor links to the report from its own domain or repository. Scammers can upload altered PDFs or use logos without permission. Multiple audits may share the same outdated scope.
## Step 8: Investigate incentives and conflicts [#step-8-investigate-incentives-and-conflicts]
Map:
* team and advisor allocations;
* vesting cliffs and unlocks;
* treasury compensation;
* market-maker loans or options;
* investor terms where disclosed;
* affiliate and influencer payments;
* related entities receiving protocol fees;
* founder sales or transfers supported by evidence.
Do not call a transfer a sale without execution evidence. Do flag an undisclosed economic relationship when independent documentation supports it.
## Step 9: Search for failures and responses [#step-9-search-for-failures-and-responses]
Look for lawsuits, regulator notices, hacks, abandoned products, failed prior tokens, censored criticism, and conflicting explanations. Distinguish allegation, filed complaint, judgment, settlement, and rumor.
How a team responds to a real failure can be more informative than a clean marketing history:
* Did it disclose promptly?
* Did it preserve evidence and publish a postmortem?
* Did claims change after the fact?
* Were affected users given a clear accounting?
* Were permissions or processes improved?
Chainalysis estimated that scams and fraud received at least $14 billion on-chain in 2025 and projected the figure could exceed $17 billion as more illicit addresses were identified; it also documented rapidly growing impersonation and AI-enabled tactics ([Chainalysis](https://www.chainalysis.com/blog/crypto-scams-2026/)). That scale makes independent verification more important, but it does not mean every pseudonymous team is fraudulent.
## An evidence graph you can reuse [#an-evidence-graph-you-can-reuse]
Create one row per claim:
| Claim | Project source | Independent source | On-chain/code evidence | Conflict | Confidence |
| ------------------- | --------------------- | -------------------- | ----------------------- | -------------------- | ------------ |
| Founder worked at X | Bio URL | Employer archive | Prior repo/talk | Dates differ | Low/med/high |
| Contract audited | Project security page | Auditor report | Commit/deployment match | New code after audit | Low/med/high |
| Treasury multisig | Docs | Explorer labels | Signers/threshold | Unknown signer | Low/med/high |
| Partner integration | Blog | Partner announcement | Live integration | Scope overstated | Low/med/high |
Confidence should attach to each claim, not the whole team. “Partially verified” is more useful than a binary legitimate/scam label.
## Red flags that justify stopping [#red-flags-that-justify-stopping]
* stolen or mismatched identities;
* guaranteed returns or risk-free language;
* pressure to transfer funds before verification;
* seed phrase, private key, or remote-access requests;
* fake audit reports or contract addresses;
* material biography claims contradicted by primary records;
* undisclosed admin powers that conflict with public claims;
* treasury or liquidity controlled by an unexplained single key;
* “partnerships” denied by the named party;
* refusal to identify the token contract or legal counterparty;
* coordinated attacks on anyone asking basic verification questions.
Investor.gov lists guarantees, risk-free claims, pressure, fake testimonials, and exaggerated credentials among classic investment-fraud warning signs ([Investor.gov](https://www.investor.gov/protect-your-investments/fraud/how-avoid-fraud/red-flags-investment-fraud-checklist)).
## How AI can help without certifying a team [#how-ai-can-help-without-certifying-a-team]
AI can collect claims, build timelines, compare biographies, search archived records, inspect repository patterns, summarize audits, and flag contradictions. Ask it to cite every material fact and distinguish absence of evidence from evidence of absence.
It cannot perform a definitive identity check from public profiles, know private arrangements, or certify motive. AI-generated biographies and deepfakes also make apparent corroboration cheaper to fabricate.
Use this prompt:
> Build an evidence graph for every team, entity, partner, audit, and control claim. Prioritize official registries, employer or partner records, auditor-hosted reports, repositories, and on-chain contracts. Mark self-referential loops, date conflicts, and unverified claims. Do not assign a universal legitimacy score or recommend an investment.
BlockMind’s agent can assist with this research, while the manual [12-step DYOR checklist](/blog/manual-dyor-checklist-12-steps) covers the rest of the asset.
If the team controls liquidity or upgrade keys, continue with the [rug-pull checker limitations matrix](/blog/crypto-rug-pull-checker-guide) rather than treating identity verification as a safety certificate.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Real people and registered companies can still act badly.
* Anonymous teams can ship secure, durable open-source software.
* Corporate and identity records differ by jurisdiction and may be private.
* Repository data can be incomplete, purchased, forked, or outsourced.
* Audit quality varies, and audited code can still contain bugs.
* Lack of online history can reflect privacy or early career rather than fraud.
The conclusion should be “verified to this confidence under these sources,” never “safe.”
## The Bottom Line [#the-bottom-line]
Verify a crypto team as a network of claims about identity, capability, accountability, and control. Seek independent evidence, inspect code and deployed permissions, read audits in scope, map incentives, and preserve unresolved conflicts. A face and a résumé are only the beginning.
This is research, not financial advice. BlockMind’s agent never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [CFTC — AI Won’t Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
* [Investor.gov — Red Flags of Investment Fraud Checklist](https://www.investor.gov/protect-your-investments/fraud/how-avoid-fraud/red-flags-investment-fraud-checklist)
* [GitHub Docs — Viewing a Project’s Contributors](https://docs.github.com/en/repositories/viewing-activity-and-data-for-your-repository/viewing-a-projects-contributors)
* [GitHub Docs — Commit Signature Verification](https://docs.github.com/en/authentication/managing-commit-signature-verification/about-commit-signature-verification)
* [ICANN — Information for RDAP Users](https://www.icann.org/en/contracted-parties/registry-operators/registration-data-access-protocol/information-for-rdap-users-31-08-2018-en)
* [Chainalysis — 2026 Crypto Crime Report: Scams](https://www.chainalysis.com/blog/crypto-scams-2026/)
---
# How to Read a Crypto Audit Report: Scope, Findings, Fixes, and Gaps
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: how to read a crypto audit
Keywords: how to read crypto audit report, smart contract audit report, crypto audit checklist, DeFi audit report, smart contract security audit
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** An audit report is evidence that a defined scope received a defined review during a defined period. It is not a safety certificate. Verify the auditor-hosted report, exact commit and files, chain and deployed addresses, methodology, unresolved findings, remediation evidence, privileged controls, and changes made after the review. If the audited code does not match the current deployment, the logo is not meaningful coverage.
***
## What does a crypto audit report actually prove? [#what-does-a-crypto-audit-report-actually-prove]
A smart contract audit report documents what one or more reviewers examined, the methods they used, the issues they identified, and the project’s response within a stated engagement.
It can support claims such as:
* A specific code revision was independently reviewed
* Reviewers used manual and automated techniques described in the report
* Particular findings were reported with stated severity
* A follow-up revision addressed or acknowledged findings
It cannot, by itself, prove:
* The current deployment matches the reviewed code
* Every vulnerability was found
* Economic design and oracle assumptions are sound
* Admin keys are secure
* Front ends, bridges, dependencies, and off-chain services are safe
* The team is honest
* Future upgrades will be audited
Ethereum.org explicitly advises against treating audits as a silver bullet and notes that they may miss bugs ([ethereum.org](https://ethereum.org/developers/docs/smart-contracts/security)). That limitation should appear in every serious audit interpretation.
**Scope and assumptions:** This guide helps non-auditors inspect public smart-contract security reports. It does not reproduce a security review, certify code, or replace qualified technical analysis. Never interact with a contract solely because it has an audit badge.
*This is research, not financial advice. BlockMind’s agent cannot transact, approve contracts, withdraw, or move funds.*
An audit review belongs inside the complete [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto); it cannot replace team, tokenomics, liquidity, governance, and deployment checks.
## The audit coverage equation [#the-audit-coverage-equation]
Think of usable coverage as an intersection:
```text
Usable audit coverage
= verified report authenticity
∩ exact code scope
∩ matching deployment
∩ resolved or accepted findings
∩ no material unaudited changes
```
If any part is missing, describe the gap. Do not collapse the result into “audited” or “not audited.”
## Step 1: verify the report is authentic [#step-1-verify-the-report-is-authentic]
Find the report on the auditor’s official domain or official repository, not only on the project’s website.
Record:
* Auditor legal or project name
* Report URL
* Publication and engagement dates
* Final versus draft status
* Named project and product
* Report version
* Auditor signature, hash, or official repository reference if provided
A logo carousel is not a report. A screenshot of an executive summary is not a report. A report uploaded only by the client deserves additional confirmation from the auditor.
## Step 2: map the exact scope [#step-2-map-the-exact-scope]
Extract the scope before reading the findings.
| Scope field | What to record |
| ----------- | ------------------------------------------------------------------- |
| Repository | Canonical URL and organization |
| Revision | Commit hash, tag, or archive hash |
| Files | Included directories/contracts |
| Exclusions | Tests, dependencies, front end, scripts, oracle, bridge, deployment |
| Chain | Network and execution environment |
| Assumptions | Trusted roles, external services, invariants |
| Duration | Review period and person-days if disclosed |
| Methodology | Manual review, tests, static analysis, fuzzing, formal methods |
Then ask what was **not** reviewed. Many important failures sit at system boundaries: price feeds, key management, deployment configuration, upgrade process, and external integrations.
## Step 3: match audited source to deployed code [#step-3-match-audited-source-to-deployed-code]
Source-code verification and formal verification are different. Ethereum.org explains that source verification checks whether published source compiles to the bytecode at an address, whereas formal verification tries to prove specified correctness properties ([ethereum.org](https://ethereum.org/developers/docs/smart-contracts/verifying/)). A “verified” explorer badge does not mean “audited” or “correct.”
For each live contract:
1. Record chain and address from current official documentation.
2. Confirm explorer source verification.
3. Determine whether it is a proxy.
4. Resolve the current implementation address.
5. Compare implementation source with the audited commit.
6. Record constructor or initializer parameters.
7. Identify linked libraries and external dependencies.
8. Check deployment date against audit date.
If exact bytecode reproduction is beyond your ability, mark the match **unverified** and seek a reproducible build or technical review. Do not infer equivalence from matching filenames.
### Proxy and upgrade checks [#proxy-and-upgrade-checks]
For an upgradeable system, capture:
* Proxy and current implementation addresses
* Proxy admin or upgrade authority
* Multisig threshold and signer disclosure, if public
* Timelock delay
* Emergency pause authority
* Upgrade events since the audit
* Whether each implementation revision was reviewed
An old audit may still cover unchanged components, but the current system needs a component-by-component map.
## Step 4: interpret finding severity correctly [#step-4-interpret-finding-severity-correctly]
Severity usually combines impact and likelihood, but each auditor defines categories differently. Read the report’s own rubric.
Common labels include:
* Critical
* High
* Medium
* Low
* Informational or note
Do not compare counts across firms without aligning definitions. One report may classify centralization or specification issues as informational; another may assign severity.
For every finding, record:
| Field | Question |
| --------------- | --------------------------------------------- |
| ID/title | Can you track it across revisions? |
| Affected code | Is it still deployed? |
| Preconditions | What must be true to exploit it? |
| Impact | What can be lost, changed, or blocked? |
| Likelihood | How reachable are the preconditions? |
| Recommendation | What did the auditor ask for? |
| Client response | Fixed, acknowledged, disputed, accepted risk? |
| Retest evidence | Did the auditor verify a revised commit? |
| Residual risk | What remains after the fix? |
“Resolved” should mean more than a client comment. Look for the fix commit and auditor retest status.
## Step 5: distinguish remediation statuses [#step-5-distinguish-remediation-statuses]
Use precise statuses:
* **Fixed and retested:** Auditor reviewed the change and marked it resolved
* **Fixed, not retested:** Client supplied a change, but no auditor confirmation is shown
* **Partially fixed:** Some exposure remains
* **Acknowledged:** Client accepts the observation without changing code
* **Risk accepted:** Client knowingly leaves the behavior
* **Disputed:** Client and auditor disagree
* **Not applicable:** The finding does not apply under a documented design assumption
* **Open:** No adequate resolution is documented
An acknowledged high-severity issue is not “closed” in the same sense as a retested fix.
## Step 6: inspect privileged and economic controls [#step-6-inspect-privileged-and-economic-controls]
Security is broader than reentrancy and arithmetic.
Ask:
* Who can mint, freeze, pause, upgrade, or rescue assets?
* Can an admin change fees, collateral factors, oracle, or withdrawal rules?
* Is there a timelock, and can it be bypassed?
* Can governance be captured by concentrated voting power?
* What happens if an oracle halts or returns a bad value?
* Are rate limits, caps, or circuit breakers enforced?
* Can external calls or tokens behave unexpectedly?
* Does the system assume honest keepers, relayers, sequencers, or validators?
The Solidity documentation maintains a current list of security considerations and known pitfalls, including external calls, gas limits, authorized proxies, and failure-handling concerns ([Solidity docs](https://docs.soliditylang.org/en/latest/security-considerations.html)). Use it as a prompt list, not evidence that a particular contract is vulnerable.
## Step 7: look for defense in depth after the audit [#step-7-look-for-defense-in-depth-after-the-audit]
An audit is one layer. Stronger evidence includes:
* Unit, integration, invariant, and fuzz tests
* Formal specifications for critical properties
* Independent reviews by different teams
* Public bug bounty with meaningful scope
* Monitoring and incident-response plan
* Timelocks, pause mechanisms, caps, and multisigs
* Reproducible builds and verified source
* Transparent disclosure of incidents and postmortems
Ethereum.org recommends combining testing approaches and independent review because no single technique finds everything ([testing guide](https://ethereum.org/developers/docs/smart-contracts/testing/)).
## Worked hypothetical: “Delta Vault” [#worked-hypothetical-delta-vault]
**Delta Vault is fictional. This is not a real audit assessment or endorsement.**
The project page displays “Audited by SecureCo.” The final SecureCo report is authentic and lists:
* Repository: `delta/vault`
* Commit: `a1b2c3`
* Scope: vault accounting and withdrawal contracts
* Excluded: oracle adapter, deployment scripts, front end
* Findings: one high, two medium, five low
Verification finds:
1. The high-severity accounting issue is fixed in `d4e5f6` and marked retested.
2. One medium finding about privileged fee changes is “acknowledged,” not fixed.
3. The live vault is a proxy deployed after the report.
4. Its current implementation maps to `h7i8j9`, three releases after the retested commit.
5. The later release changed the excluded oracle adapter and withdrawal queue.
6. A two-of-three multisig can upgrade immediately; there is no timelock.
Responsible conclusion:
> The audit is authentic and provides evidence for the scoped contracts at the retested revision. The current deployment contains material later changes and an excluded oracle component, so full deployment coverage is unverified. A medium privileged-control issue remains acknowledged, and immediate two-of-three upgrades create residual admin risk.
“Audited” would have hidden every important detail.
## An audit report reading template [#an-audit-report-reading-template]
```text
REPORT IDENTITY
- Auditor-hosted final URL:
- Dates and version:
- Project/product:
SCOPE
- Repository:
- Commit/tag/hash:
- Included files:
- Excluded components:
- Methodology and timebox:
- Stated assumptions:
DEPLOYMENT MATCH
- Chain and addresses:
- Proxy / implementation:
- Verified source:
- Audited revision match:
- Post-audit changes:
FINDINGS
- Critical/high/medium/low counts under auditor rubric:
- Open or accepted findings:
- Fixed and retested findings:
- Residual risks:
PRIVILEGED CONTROLS
- Upgrade authority:
- Multisig threshold:
- Timelock:
- Pause/freeze/mint/rescue powers:
DEFENSE IN DEPTH
- Tests/formal methods:
- Bug bounty:
- Monitoring/incident response:
- Additional audits:
CONCLUSION
- What the audit supports:
- What it does not support:
- Unresolved verification tasks:
```
## Audit red flags [#audit-red-flags]
* No full report or auditor-side confirmation
* Report marked draft or preliminary
* Scope has no commit, tag, files, or hash
* Audit predates a major deployment or upgrade
* Current addresses are missing from official docs
* Explorer source is unverified or only partially matched
* Proxy implementation has changed since review
* Findings are called “resolved” based only on client statements
* High-impact observations are “acknowledged” without residual-risk explanation
* Auditor’s disclaimers and exclusions are omitted from marketing
* Audit covers a token contract while the value-bearing protocol is unaudited
* Multiple audits all review the same old revision
* No incident plan, monitoring, or bug bounty for a high-value protocol
## Limitations and counterevidence [#limitations-and-counterevidence]
* A clean report may reflect good code, a narrow scope, a short timebox, or missed issues.
* A report with many findings can indicate weak initial code or a thorough review followed by strong remediation.
* Public readers may be unable to reproduce builds or assess exploitability.
* An auditor’s reputation is relevant but not a guarantee.
* Formal verification proves specified properties; a bad or incomplete specification can miss system risk.
* Bug bounties and multiple audits improve coverage but cannot eliminate risk.
Do not rank projects by finding count alone.
## Using BlockMind to organize an audit review [#using-blockmind-to-organize-an-audit-review]
A BlockMind agent can locate the auditor-hosted report, extract scope and findings, compare current documentation, and visit explorer and repository pages through its [browser capability](/docs/capabilities/browser). It can save the scope map and unresolved checks in your Notebook.
Ask:
> “Read this final audit report. Extract the exact scope, commit, exclusions, severity rubric, every non-retested or accepted finding, and all privileged-control assumptions. Then compare the current official deployment addresses and implementation revision. Mark anything you cannot verify as unknown.”
AI extraction can be wrong. Use the [AI Crypto Analysis Verification Protocol](/blog/how-to-verify-ai-crypto-analysis), and include audit evidence inside the broader [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps).
## The Bottom Line [#the-bottom-line]
A smart contract audit is a scoped review, not a transferable seal of safety. Verify the report’s authenticity, exact revision, included and excluded components, deployment match, remediation status, admin powers, and every post-audit change.
The right conclusion is specific: which code received which review, which findings remain, and whether that evidence still reaches the contracts users interact with today.
## Sources [#sources]
* [Ethereum.org — Smart Contract Security](https://ethereum.org/developers/docs/smart-contracts/security)
* [Ethereum.org — Testing Smart Contracts](https://ethereum.org/developers/docs/smart-contracts/testing/)
* [Ethereum.org — Verifying Smart Contracts](https://ethereum.org/developers/docs/smart-contracts/verifying/)
* [Solidity Documentation — Security Considerations](https://docs.soliditylang.org/en/latest/security-considerations.html)
* [ERC-7512 — Onchain Representation for Audits and Its Limitations](https://eips.ethereum.org/EIPS/eip-7512)
* [Ethereum Foundation Security — Published Audit Reports](https://security.ethereum.org/audits/)
---
# How to Read Crypto Social Sentiment Without Chasing Hype
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: problem-solving
Primary keyword: how to read crypto social sentiment
Keywords: how to read crypto social sentiment, crypto sentiment analysis, social sentiment crypto, crypto social volume, X sentiment crypto
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Do not reduce crypto social sentiment to “72% bullish.” Read five dimensions together: how much attention changed, what the tone is, whether activity looks authentic, how diverse the sources are, and whether the change persists. Then align sentiment with price, liquidity, on-chain activity, and event timestamps. Sentiment can lead price, follow price, or be manufactured; the direction must be tested, not assumed.
Social data matters because crypto narratives, communities, and scams all spread through public networks. It is also one of the easiest datasets to misread. A feed is personalized, language models struggle with slang and sarcasm, and coordinated accounts can create the appearance of consensus.
Sentiment belongs in a multi-signal process such as the [crypto market analysis guide](/blog/crypto-market-analysis-guide), not in an isolated buy-or-sell rule.
The workflow comparison in [AI crypto analysis vs X/Twitter](/blog/ai-crypto-analysis-vs-twitter) shows how to turn a social claim into a source-verification task.
## What crypto sentiment analysis actually measures [#what-crypto-sentiment-analysis-actually-measures]
Sentiment tools usually combine several different variables:
* **Mention volume:** how often an asset or narrative appears.
* **Unique authors:** how many accounts participated.
* **Engagement:** replies, reposts, likes, views, or forum votes.
* **Tone:** positive, negative, neutral, fearful, excited, angry, or uncertain language.
* **Topic:** what people are discussing—price, product, exploit, listing, governance, or memes.
* **Influence:** estimated reach or network position of the authors.
* **Velocity:** how quickly the measures change versus a baseline.
Each answers a different question. Mention volume measures attention, not approval. Positive tone measures language, not truth. Engagement measures distribution, not independent belief.
## The five-factor reading framework [#the-five-factor-reading-framework]
### 1. Attention: Is the conversation actually growing? [#1-attention-is-the-conversation-actually-growing]
Compare current mention volume and unique authors with the asset’s own baseline, not with Bitcoin or a much larger project. Useful views include:
* current hour/day versus trailing median;
* unique-author growth versus raw-post growth;
* search interest versus social mentions;
* share of category conversation rather than absolute count.
A spike dominated by repeated posts from the same accounts is different from broad discovery.
### 2. Tone: What emotion and claim changed? [#2-tone-what-emotion-and-claim-changed]
“Bullish” is too coarse. Separate:
* product optimism;
* price excitement;
* fear or security concern;
* anger at governance;
* skepticism about evidence;
* urgency or FOMO;
* neutral news repetition.
Read representative posts from each cluster. A negative spike caused by researchers responsibly disclosing an exploit differs from generic despair after a market decline.
### 3. Authenticity: Does the activity look coordinated? [#3-authenticity-does-the-activity-look-coordinated]
Watch for:
* newly created or dormant accounts activating together;
* identical or lightly rewritten phrases;
* synchronized posting and reposting;
* repeated links, referral codes, or contract addresses;
* engagement rings that interact mostly with one another;
* high post volume with little genuine reply conversation;
* sudden follower or engagement discontinuities;
* undisclosed promotion.
These are indicators, not proof of bots or payment. X’s platform-manipulation policy prohibits coordinated inauthentic activity and artificial engagement, confirming that platform metrics can be deliberately manipulated ([X](https://help.x.com/en/rules-and-policies/platform-manipulation)).
### 4. Source diversity: Is this consensus or one claim copied widely? [#4-source-diversity-is-this-consensus-or-one-claim-copied-widely]
Cluster posts by the underlying source. Twenty outlets and hundreds of accounts may all trace to one project announcement or anonymous screenshot. Count independent evidence chains, not URLs.
Compare:
* official project and developer accounts;
* independent technical researchers;
* known investors and promoters;
* news organizations;
* community members;
* anonymous accounts;
* other platforms and languages.
Disagreement between informed groups is often more useful than an aggregate score.
### 5. Persistence: Did the change survive the first reaction? [#5-persistence-did-the-change-survive-the-first-reaction]
Measure whether attention and tone remain elevated after several observation windows. Short spikes often reflect a listing, rumor, campaign, or price candle. Persistent, diverse discussion around usage or governance may be more informative, but still needs fundamental verification.
## Add the missing sixth step: lead-lag direction [#add-the-missing-sixth-step-lead-lag-direction]
Plot sentiment measures beside price, volume, liquidity, and the actual event time. Ask:
1. Did social attention rise before, during, or after the price move?
2. Did tone change before attention, or did price excitement change both?
3. Was an external event published first?
4. Does the relationship persist across more than one episode?
5. Does the result survive alternative windows and lag choices?
Academic findings are mixed and asset-specific. A 2025 study using 66,582 Reddit posts about Bitcoin and 23,231 about Ethereum found Bitcoin returns more sensitive to negative sentiment, Ethereum returns unaffected by the tested sentiment types, and a bidirectional relationship between markets and sentiment ([Blockchain: Research and Applications](https://doi.org/10.1016/j.bcra.2025.100375)). A separate 2025 forecasting study found that textual features could improve models in its experimental setting, but that does not establish a universal live trading rule ([International Journal of Forecasting](https://doi.org/10.1016/j.ijforecast.2025.02.007)).
The responsible conclusion is that sentiment may add information under some methods and regimes—not that a positive score predicts a rally.
## A worked interpretation pattern [#a-worked-interpretation-pattern]
Imagine a token’s mentions rise fivefold while the sentiment score becomes strongly positive.
Do not conclude “bullish.” Ask:
### Attention [#attention]
Did unique authors also rise, or did a small group post repeatedly?
### Tone [#tone]
Are people discussing a shipped product, or only the price and referral rewards?
### Authenticity [#authenticity]
Are phrases, timestamps, and engagement networks unusually similar?
### Diversity [#diversity]
Do independent developers, users, and researchers participate, or does everything trace to the project campaign?
### Persistence [#persistence]
Does discussion continue after 24–72 hours, and does substantive content replace promotional repetition?
### Confirmation [#confirmation]
Do liquidity, holder distribution, contract activity, fees, and official releases support the narrative? If social attention rises while liquidity and real usage remain flat, the gap is itself a risk signal.
## Sentiment patterns and what they may mean [#sentiment-patterns-and-what-they-may-mean]
| Pattern | Plausible reading | What to check next |
| ----------------------------------------------- | -------------------------------- | ----------------------------------------------------------- |
| Attention up, tone neutral | News discovery | Primary event source and topic clusters |
| Attention up, tone euphoric, price already up | Reaction/FOMO | Event time, liquidity, promotion, late entrants |
| Attention up, tone negative, price flat | Emerging concern | Security disclosure, governance dispute, source credibility |
| Positive tone, few unique authors | Coordinated campaign possible | Author network and repeated text |
| Diverse technical discussion, modest engagement | Potentially substantive interest | Repositories, docs, and on-chain usage |
| Sentiment improves while liquidity falls | Fragile divergence | Spread, depth, holder transfers |
| Price falls before sentiment turns negative | Sentiment likely reacting | Earlier market or on-chain catalyst |
None of these patterns is a trade signal by itself.
## How manipulation changes the reading [#how-manipulation-changes-the-reading]
The CFTC warns that pump-and-dump organizers use social media and messaging channels to hype thinly traded tokens, and advises against buying from a single social tip or sudden spike ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)). The FTC reported $2.1 billion in reported losses from scams that began on social media in 2025, with investment scams accounting for $1.1 billion of that total ([FTC](https://www.ftc.gov/news-events/news/press-releases/2026/04/new-ftc-data-show-people-have-lost-billions-social-media-scams)).
Those figures describe reported consumer fraud across social platforms, not the error rate of every crypto post. They do establish a high bar for turning online excitement into financial action.
When the activity includes coordinated entries, urgency, and thin liquidity, use the specific [crypto pump-and-dump signs](/blog/crypto-pump-and-dump-signs) checklist.
Red flags include:
* guaranteed returns or “zero risk”;
* urgency and countdowns;
* private groups promising coordinated entries;
* requests to send crypto to unlock profits;
* fake celebrity or support accounts;
* screenshots without verifiable transaction or account context;
* a token contract distributed only through replies or DMs;
* promoters who block questions about liquidity, ownership, or compensation.
## How AI can help—and fail [#how-ai-can-helpand-fail]
AI can cluster topics, detect repeated language, compare time windows, translate multilingual posts, summarize disagreement, and join social signals with market or on-chain evidence. It can make a noisy dataset inspectable.
Ask it to return:
* collection source and query;
* asset-resolution rules;
* sampling period and timezone;
* number of posts and unique authors;
* language coverage;
* bot/duplication filtering method;
* tone taxonomy and validation method;
* representative examples;
* event and price alignment;
* uncertainty and missing channels.
AI can fail through sarcasm errors, dialect bias, ticker ambiguity, sampling restrictions, bot misclassification, and hindsight storytelling. A summary of a personalized feed is not population sentiment. BlockMind’s research capability can include X/Twitter alongside market and web research, but it should be used as one evidence layer: [Research](/docs/capabilities/research).
## A daily sentiment review template [#a-daily-sentiment-review-template]
```text
Asset and contract:
Window and timezone:
Platforms and query:
Attention change vs baseline:
Unique-author change:
Tone by topic:
Source diversity:
Authenticity concerns:
Event timestamp:
Price/liquidity/on-chain confirmation:
Lead-lag interpretation:
Counterevidence:
What remains unknown:
Next review condition:
```
Using the same template matters more than chasing a proprietary score. Consistency reveals when a relationship changes.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Platform APIs expose samples, not necessarily the full conversation.
* Deleted, private, and closed-channel messages are missing.
* Language and sarcasm classification is imperfect.
* High-follower accounts can dominate weighted scores.
* A bot filter can remove real coordinated communities or retain sophisticated automation.
* Price influences sentiment, creating reverse causality.
* Backtested relationships can decay after platform, market, or participant changes.
Qualitative reading still has value: a domain expert may notice a credible technical concern before any aggregate metric moves. Do not let a positive score erase specific counterevidence.
## The Bottom Line [#the-bottom-line]
Read social sentiment as a structured observation of attention and narrative, not a forecast. Measure attention, tone, authenticity, diversity, persistence, and lead-lag direction. Then cross-check the story against primary sources, liquidity, on-chain activity, and holder behavior.
This is research, not financial advice. BlockMind’s agent never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [CFTC — Beware Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
* [FTC — New Data Show People Have Lost Billions to Social Media Scams](https://www.ftc.gov/news-events/news/press-releases/2026/04/new-ftc-data-show-people-have-lost-billions-social-media-scams)
* [X — Platform Manipulation and Spam Policy](https://help.x.com/en/rules-and-policies/platform-manipulation)
* [Blockchain: Research and Applications — Virtual Influence, Real Impact](https://doi.org/10.1016/j.bcra.2025.100375)
* [International Journal of Forecasting — Deep Learning and NLP in Cryptocurrency Forecasting](https://doi.org/10.1016/j.ijforecast.2025.02.007)
---
# How to Read a Crypto Whitepaper: A Claim-by-Claim Method
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: how to read a crypto whitepaper
Keywords: how to read a crypto whitepaper, crypto whitepaper analysis, whitepaper red flags, cryptocurrency white paper checklist, DYOR whitepaper
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Do not read a crypto whitepaper as proof. Read it as the project’s organized set of claims. In four passes—identity, mechanism, economics, and verification—turn every material claim into a source, test, owner, date, and failure condition. The best outcome is not “the paper sounds convincing”; it is “I know which claims are verified, contradicted, stale, or still unknown.”
***
## How should you read a crypto whitepaper? [#how-should-you-read-a-crypto-whitepaper]
Read a crypto whitepaper in four passes:
1. **Identity and scope:** Who wrote it, when, for which version, and what rights are described?
2. **Mechanism:** What problem, system, trust assumptions, and technical tradeoffs does it claim?
3. **Economics:** Who receives tokens, what creates supply and demand, and which actors can change the rules?
4. **Verification:** Which claims match code, contracts, governance, market data, and current documentation?
The paper is a starting document, not a certificate of truth. Ethereum’s own whitepaper page warns that its original 2014 paper no longer reflects Ethereum today after years of upgrades ([ethereum.org](https://ethereum.org/whitepaper/)). That is the clearest possible reminder that even a historically important whitepaper can become stale.
**Scope and assumptions:** This method is for protocol, network, application, and token whitepapers. It is not a legal determination about whether an asset is a security or whether a disclosure satisfies a jurisdiction’s rules. Where regulatory status or rights matter, consult qualified legal sources.
*This is research, not financial advice. BlockMind does not tell you what to buy or sell, and its agent cannot touch your funds.*
This claim-by-claim review is one layer of the broader [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto), which also covers team, market, security, holder, and risk evidence.
## What a whitepaper can and cannot establish [#what-a-whitepaper-can-and-cannot-establish]
A whitepaper can establish what its authors chose to claim at a particular time. It can explain a proposed design, token role, distribution, governance process, risk model, and roadmap.
By itself, it cannot establish that:
* The named team is real or accountable
* The published code implements the paper
* The deployed contracts match the reviewed code
* The token allocation remains unchanged
* Stated partnerships are acknowledged by the other party
* Users, revenue, or liquidity exist
* An audit covers the current deployment
* The roadmap will be completed
* The token will appreciate
The CFTC advises readers to understand what rights attach to a token, how funds will be used, and whether affiliates can be verified; it also warns against promises or guarantees of future value ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)). That is a practical reading mandate: claims about rights, people, money, and value deserve independent evidence.
## Pass 1: establish identity, version, and rights [#pass-1-establish-identity-version-and-rights]
Before reading the technical sections, capture the document’s identity.
### Record the document fingerprint [#record-the-document-fingerprint]
* Exact title
* Canonical URL
* Publication date and revision date
* Version number or commit hash, if present
* Authors and issuing entity
* Archived copy or local hash for later comparison
* Network and product version described
A PDF forwarded through a chat is not a canonical source. Find the project’s official site or repository and preserve the retrieval date.
### Ask who is accountable [#ask-who-is-accountable]
Look for a legal entity, contributors, governing body, conflict disclosures, and contact method. Anonymous contributors do not prove fraud, but anonymity changes the available accountability evidence. Compensate with stronger code, governance, treasury, and deployment verification.
### Extract the rights—not the slogans [#extract-the-rightsnot-the-slogans]
Write down what holding the token actually permits:
* Governance proposal or voting rights
* Fee payment or gas use
* Access to a product or service
* Staking or protocol-security role
* Claims, redemption, or reserve rights
* No enforceable right at all
Do not translate “community-owned” into voting power without checking delegation, quorum, proposal thresholds, privileged roles, and actual token distribution.
The EU Markets in Crypto-Assets regulation is a legal disclosure framework, not an investment checklist. Still, its required categories are useful prompts: issuer, project, offer, attached rights and obligations, underlying technology, and risks; asset-referenced tokens require additional reserve information ([EUR-Lex, Regulation 2023/1114](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32023R1114)). Use these categories to spot omissions, not to assume compliance.
## Pass 2: reconstruct the mechanism [#pass-2-reconstruct-the-mechanism]
The mechanism pass asks whether the system can be explained without marketing adjectives.
### Write the problem in one sentence [#write-the-problem-in-one-sentence]
Use this form:
> “For \[user], the project claims to reduce \[specific cost or constraint] by \[mechanism], compared with \[current alternative].”
If you cannot fill the blanks after reading the paper, the document has not explained the problem clearly.
### Draw the actor-and-trust map [#draw-the-actor-and-trust-map]
List every actor that can change an outcome:
* Users
* Validators, sequencers, miners, or operators
* Tokenholders and delegates
* Foundation or company
* Multisig signers
* Oracle providers
* Bridges or custodians
* Upgrade administrators
* External service providers
For each actor, ask:
1. What can it do?
2. What prevents abuse?
3. What happens if it disappears?
4. Can users exit before a harmful change?
A “decentralized” label is less informative than an explicit map of permissions and failure modes.
### Translate architecture into testable invariants [#translate-architecture-into-testable-invariants]
An invariant is a property the system claims should remain true. Examples:
* Total supply cannot exceed a stated cap
* Withdrawals require a valid proof
* A privileged change has a minimum delay
* Collateral value must remain above liabilities
* No single signer can upgrade the system
Then identify where each invariant is enforced: consensus rules, smart contracts, governance process, legal promise, or operational policy. These are not equivalent guarantees.
Bitcoin’s original paper is valuable because it lays out a specific mechanism—transactions, timestamp server, proof of work, incentives, and simplified verification—rather than only a market vision ([Bitcoin.org](https://bitcoin.org/en/bitcoin-paper)). Use that mechanism-level specificity as a benchmark, while still verifying the live implementation separately.
## Pass 3: rebuild the economics [#pass-3-rebuild-the-economics]
The economics pass should produce a supply-and-incentive table, not a paragraph of impressions.
### Reconcile the supplies [#reconcile-the-supplies]
Record:
* Initial or genesis supply
* Current circulating supply
* Total supply
* Maximum supply, if one exists
* Mint and burn authorities
* Emission formula
* Locked allocations and [unlock schedule](/blog/token-unlocks-vesting-guide)
Check the paper against current explorer data, official supply endpoints, governance changes, and the project’s latest docs. If two aggregators disagree, do not average them; trace the definition and source.
### Build the allocation table [#build-the-allocation-table]
| Allocation | Share or tokens | Recipient | Lock/vesting | Control before unlock | Source |
| -------------------- | --------------: | --------- | ------------ | --------------------- | ------ |
| Team | | | | | |
| Investors | | | | | |
| Treasury | | | | | |
| Community incentives | | | | | |
| Foundation/ecosystem | | | | | |
The “control before unlock” column matters. Tokens can be contract-locked, contractually restricted, held in a normal wallet, delegated, lent to a market maker, or transferable through ownership of a vesting vehicle. The economic effect differs.
### Follow the incentive loop [#follow-the-incentive-loop]
For every reward, ask where it comes from:
* Fees paid by users
* New token issuance
* Treasury subsidy
* External revenue
* Another participant’s loss
Then ask what recipients are expected to do with it. Yield funded by issuance can be useful for bootstrapping, but it is not the same as yield funded by external demand.
Use [Market Cap vs FDV](/blog/market-cap-vs-fdv-crypto) to distinguish current public float from a full-supply scenario, and [How to Research Crypto Tokenomics](/blog/how-to-research-crypto-tokenomics) for the broader supply framework.
## Pass 4: turn claims into a verification ledger [#pass-4-turn-claims-into-a-verification-ledger]
This is the step most whitepaper reviews skip.
Copy every material claim into this ledger:
| ID | Claim | Claim type | Primary evidence needed | Status | Failure condition | Checked |
| ---- | ------------------------------- | ---------- | --------------------------------- | ------- | ---------------------- | ------- |
| C-01 | “Supply is capped” | Protocol | Code + live mint state | Unknown | Mint path exceeds cap | Date |
| C-02 | “Governance controls upgrades” | Governance | Contracts + docs + proposals | Unknown | Admin can bypass vote | Date |
| C-03 | “Partner X integrates protocol” | Commercial | Partner’s own announcement | Unknown | Only project claims it | Date |
| C-04 | “Audit completed” | Security | Final report + deployed addresses | Unknown | Scope/version mismatch | Date |
Use five statuses:
* **Verified:** Primary evidence directly supports the claim
* **Partly verified:** Some elements match, others remain open
* **Unverified:** No adequate independent evidence found
* **Contradicted:** Better evidence conflicts with it
* **Stale:** It was accurate for an older version or date
“Not verified” does not mean false. It means the research state is unresolved.
### Verification order by claim type [#verification-order-by-claim-type]
| Claim type | Best first source |
| ------------- | --------------------------------------------------------------------- |
| Code behavior | Verified deployed source, repository release, technical specification |
| Token supply | On-chain mint state, protocol rules, official current disclosure |
| Audit | Auditor-hosted final report and exact in-scope commit/addresses |
| Partnership | Counterparty’s official site or filing |
| Users/revenue | Reproducible query and disclosed methodology |
| Governance | Live contracts, proposal system, voting rules, executed history |
| Team identity | Government/company registers, prior repositories, acknowledged work |
## Worked hypothetical: “Harbor Protocol” [#worked-hypothetical-harbor-protocol]
**This is a fictional example. It is not a real project or investment analysis.**
Harbor’s paper claims: “A decentralized lending market governed entirely by HBR holders, with a fixed one-billion-token supply and audited contracts.”
### Pass 1 result [#pass-1-result]
The PDF is version 1.2 from 18 months ago. An entity is named, but the paper has no revision log. “Governed entirely” is not defined.
### Pass 2 result [#pass-2-result]
The actor map shows borrowers, lenders, an oracle, a three-of-five upgrade multisig, and token governance. The multisig can upgrade without an on-chain token vote. The decentralization claim is therefore contradicted as written, even if governance controls other parameters.
### Pass 3 result [#pass-3-result]
The paper says one billion maximum supply. The verified token contract exposes no mint function, supporting the cap for that contract. However, 70% is non-circulating, and the paper gives no dated vesting table. Supply cap is verified; near-term float is unresolved.
### Pass 4 result [#pass-4-result]
An auditor’s final report exists, but it covers commit `abc123`. The deployed bytecode maps to a later release with changes to oracle handling. “Audited contracts” becomes **partly verified**, pending a diff and follow-up review.
The final note is not “good project” or “bad project.” It is:
> Fixed supply is supported for the identified contract. Governance language overstates tokenholder control because a multisig retains direct upgrade authority. Audit coverage does not yet match the deployed revision. Vesting remains unverified.
That note is useful because every sentence can be retested.
## Crypto whitepaper red flags that deserve follow-up [#crypto-whitepaper-red-flags-that-deserve-follow-up]
Red flags are prompts for verification, not automatic verdicts:
* No canonical version, date, authors, or revision history
* Guaranteed-return or future-price language
* Technical complexity without threat model or tradeoffs
* “Decentralized” without admin-key or upgrade disclosure
* Token utility described only as appreciation, rewards, or vague governance
* Allocation percentages that do not total correctly
* Supply definitions that mix circulating, unlocked, total, and maximum
* Roadmap milestones without owners, dependencies, or acceptance tests
* Partners named without reciprocal confirmation
* Audit logo without report, scope, commit, dates, or remediation status
* Claimed open source with no build instructions or deployed-code match
* Risk section that only describes market volatility
The absence of a risk section is especially informative. MiCA’s disclosure structure explicitly includes risk information and requires clear, non-misleading presentation for covered whitepapers; regardless of jurisdiction, a project that explains upside but not failure modes is giving you an incomplete model.
## Questions to ask after reading [#questions-to-ask-after-reading]
1. Which three claims matter most to whether the system works?
2. Which one is supported only by the issuer?
3. What live evidence has changed since publication?
4. Who can upgrade, pause, mint, freeze, or move treasury assets?
5. What rights does the token confer, and where are they enforced?
6. What is the next material supply change?
7. Does the audit match the current deployment?
8. What observation would disprove the adoption claim?
9. What did the paper omit that a competitor discloses?
10. Can another researcher reproduce your conclusion from your sources?
Pair this with [How to Read a Crypto Audit Report](/blog/how-to-read-crypto-audit-report), [Crypto GitHub Developer Activity](/blog/crypto-github-developer-activity), and the [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps).
## Limitations and counterevidence [#limitations-and-counterevidence]
* Some valuable protocols began with informal or incomplete papers and improved through open development.
* A polished disclosure can still describe a weak or malicious system.
* Code verification requires technical skill; absence of reader expertise is not evidence of safety.
* Regulatory disclosure requirements vary by asset and jurisdiction.
* Private commercial agreements may prevent full partnership or allocation disclosure.
* A whitepaper can be accurate at publication and stale today.
Treat document quality as one evidence layer. It cannot replace live technical, economic, security, and governance checks.
## Using AI without outsourcing the reading [#using-ai-without-outsourcing-the-reading]
Your BlockMind agent can extract claims, compare versions, visit live documentation, and build a verification ledger. Its [browser capability](/docs/capabilities/browser) can preserve current pages as screenshots or PDFs, while its [research capabilities](/docs/capabilities/research) can cross-check public sources.
Ask for outputs that expose evidence:
> “Extract the ten most material claims from this whitepaper. For each, quote no more than needed, identify the primary source required to verify it, mark the current evidence status, and list what would contradict it.”
Then manually open the decisive sources. Follow the [AI crypto analysis verification protocol](/blog/how-to-verify-ai-crypto-analysis) before relying on the result.
## The Bottom Line [#the-bottom-line]
A crypto whitepaper is not a conclusion. It is an input. Read it four times: identify the document and rights, reconstruct the mechanism, rebuild the economics, and verify the material claims against current primary evidence.
When you finish, you should have fewer impressions and more statuses: verified, partly verified, unverified, contradicted, or stale. That is what turns reading into due diligence.
## Sources [#sources]
* [Bitcoin.org — Bitcoin: A Peer-to-Peer Electronic Cash System](https://bitcoin.org/en/bitcoin-paper)
* [Ethereum.org — Ethereum Whitepaper and Current-Version Warning](https://ethereum.org/whitepaper/)
* [EUR-Lex — Regulation (EU) 2023/1114 on Markets in Crypto-Assets](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32023R1114)
* [EUR-Lex — Implementing Regulation (EU) 2024/2984 Whitepaper Templates](https://eur-lex.europa.eu/legal-content/en/TXT/?uri=CELEX%3A32024R2984)
* [CFTC — Use Caution When Buying Digital Coins or Tokens](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)
---
# Crypto Tokenomics Explained: How to Research Tokenomics Without a Finance Degree
Published: 2026-03-10
Updated: 2026-07-10
Review by: 2027-07-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: how to research crypto tokenomics
Keywords: crypto tokenomics, tokenomics analysis, token supply
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Tokenomics is the system behind a token's supply, distribution, incentives, and utility: how many tokens exist, who gets them, when they unlock, what changes supply, and why anyone should hold or use the token. If you can check supply, unlocks, utility, and emissions, you can understand most tokenomics without a finance degree or a 60-page whitepaper.
***
Most people overcomplicate tokenomics. You do not need to build a spreadsheet, learn corporate finance, or read every governance proposal to get the basics right.
You need to answer a simpler question: **what could increase selling pressure, what could reduce supply, and why should anyone hold the token at all?**
That is the core of tokenomics research.
For the broader research process, read [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research). Tokenomics is only one layer; team credibility, liquidity, audits, holders, and actual usage can contradict an attractive supply schedule.
## What is crypto tokenomics? [#what-is-crypto-tokenomics]
Crypto tokenomics is the system that controls a token's supply, distribution, incentives, and utility.
In plain English, tokenomics tells you five things:
1. **How many tokens exist now**
2. **How many tokens could exist later**
3. **Who owns a big share of them**
4. **When more tokens enter the market**
5. **Why the token has any reason to be held or used**
If price is what the market is paying today, tokenomics is part of the reason the market pays that price at all.
This matters because two tokens can look identical on a chart while having completely different supply pressure under the surface. One might have most of its supply already circulating. The other might have massive investor unlocks coming over the next 12 months.
That is why tokenomics research is less about memorizing jargon and more about understanding future pressure.
## How do you read tokenomics without reading a whole whitepaper? [#how-do-you-read-tokenomics-without-reading-a-whole-whitepaper]
Start with four checks: supply, unlocks, utility, and emissions.
If you only remember one framework, use this:
### 1. Check the supply numbers [#1-check-the-supply-numbers]
Look for these three numbers first:
* **Circulating supply:** tokens currently available in the market
* **Total supply:** tokens created minus any permanently burned tokens
* **Max supply:** the hard cap, if one exists
These numbers tell you whether dilution risk is small or still ahead of you.
A token with 90% of supply already circulating is a different bet from a token with only 15% circulating. The second token may still have years of unlocks ahead.
### 2. Check the vesting and unlock schedule [#2-check-the-vesting-and-unlock-schedule]
A vesting schedule tells you when team, investor, advisor, or treasury tokens become sellable.
This is one of the most important parts of tokenomics because unlocks create potential supply pressure. If a large allocation unlocks every month, early holders may keep selling into the market even if the project itself is improving.
### 3. Check what causes inflation or deflation [#3-check-what-causes-inflation-or-deflation]
Some tokens add new supply through emissions. Others reduce supply through burns. Some do both at the same time.
The key question is not “is inflation bad?” The key question is **who receives new supply, how fast, and does demand grow faster than supply?**
### 4. Check whether the token actually does anything [#4-check-whether-the-token-actually-does-anything]
Good token utility creates a reason to buy, hold, stake, spend, or govern with the token. Weak utility means the token may just exist because the project wanted a token.
If you cannot explain the token's purpose in one or two sentences, that is already a warning sign.
## What do circulating supply, total supply, and max supply actually mean? [#what-do-circulating-supply-total-supply-and-max-supply-actually-mean]
Circulating supply is what the market can realistically trade now, total supply is how many tokens exist, and max supply is the ceiling if the protocol has one.
A quick way to think about it:
* **Circulating supply** = what is out in the wild
* **Total supply** = what exists overall
* **Max supply** = the upper limit, if there is one
### Simple example [#simple-example]
Imagine a token with:
* 100 million **max supply**
* 60 million **total supply**
* 18 million **circulating supply**
That means 18 million tokens are already trading, 42 million exist but are locked or otherwise not circulating, and another 40 million could still be created later.
That should immediately raise follow-up questions:
* Who controls the locked supply?
* When does it unlock?
* What conditions create the remaining 40 million?
This is why “low market cap” alone can be misleading. A token can look cheap on circulating market cap while hiding much larger fully diluted supply.
## Why do vesting schedules matter so much? [#why-do-vesting-schedules-matter-so-much]
Vesting schedules matter because locked tokens often become future sellers.
Projects usually allocate tokens to founders, employees, investors, ecosystem funds, and community programs. That is normal. The important part is **how fast those allocations unlock**.
Here is the practical issue: if insiders received tokens far below current market price, they may be willing sellers long before retail buyers understand the unlock calendar.
When you review a vesting schedule, ask:
* Is there a **cliff** before tokens unlock?
* Are unlocks **monthly, quarterly, or yearly**?
* How large is the **team + investor allocation**?
* Is the schedule transparent and easy to verify?
* Are major unlocks clustered around the same period?
As a rough rule, a token with aggressive insider allocations and frequent unlocks deserves more skepticism than one with long, gradual vesting.
This is also why tokenomics should never be analyzed without distribution. A “great product” can still be a weak token if the structure favors early holders over later buyers.
## What is the difference between inflationary and deflationary tokenomics? [#what-is-the-difference-between-inflationary-and-deflationary-tokenomics]
Inflationary tokenomics increases supply over time, while deflationary tokenomics reduces supply over time.
But the real world is messier than those labels suggest.
Many networks are **disinflationary**, not strictly deflationary. That means supply is still growing, but it grows more slowly over time. Others can switch between inflation and deflation depending on network activity.
### Bitcoin: fixed cap with declining issuance [#bitcoin-fixed-cap-with-declining-issuance]
Bitcoin is the cleanest example of supply-driven tokenomics.
Bitcoin has a hard cap of **21 million BTC**, and its issuance falls every **210,000 blocks**, or roughly every four years. The most recent halving happened on **April 20, 2024**, cutting the block reward to **3.125 BTC** ([source](https://www.investopedia.com/bitcoin-halving-4843769)). Bitcoin is still issuing new coins, but at a slower rate than before.
That makes Bitcoin **disinflationary** in issuance terms: new supply keeps coming, but the flow keeps shrinking.
### Ethereum: no fixed cap, but a burn mechanism [#ethereum-no-fixed-cap-but-a-burn-mechanism]
Ethereum is more dynamic.
EIP-1559 introduced a pricing system where the **base fee is burned** ([source](https://eips.ethereum.org/EIPS/eip-1559)). In plain English, part of the transaction fee is permanently removed from supply. Ethereum therefore does not have a fixed maximum supply like Bitcoin, but network usage can offset or even outpace new issuance in some periods.
That is why ETH is better understood as a token with **activity-dependent supply pressure**. When usage rises, more ETH can be burned. When usage falls, net supply can grow faster.
The lesson is simple: do not stop at “capped” or “uncapped.” Ask what actually changes circulating supply in practice.
## What should you look for in a token emission schedule? [#what-should-you-look-for-in-a-token-emission-schedule]
An emission schedule tells you how new tokens enter the market over time.
This matters most for staking, mining, liquidity incentives, ecosystem rewards, and treasury distributions.
A healthy research question is: **who gets the new tokens, and what behavior is the protocol trying to encourage?**
Common emission destinations include:
* Validators or miners securing the network
* Stakers locking tokens
* Liquidity providers
* Developer or ecosystem funds
* Community rewards
High emissions are not automatically bad. They can make sense early if the protocol is bootstrapping security or liquidity. The problem starts when emissions stay high long after growth slows, because then the token can become a machine that constantly creates sell pressure.
If a token offers very high yield, always ask where that yield comes from. In many cases, the answer is simply “more token issuance.”
## What are the main types of token utility? [#what-are-the-main-types-of-token-utility]
Token utility is the reason a token exists inside a system instead of just being attached to it.
The most common utility types are:
* **Governance:** holders vote on proposals
* **Gas or transaction fees:** the token pays for network use
* **Staking/security:** the token helps secure the network
* **Access:** holding the token unlocks features or products
* **Incentives:** the token rewards user behavior
* **Collateral or settlement:** the token is used in financial activity inside the protocol
This is where a lot of projects get weak. Saying a token has “governance utility” is not enough if few holders vote and governance changes very little.
Strong utility usually means the token is tied to real usage, real demand, or real coordination inside the network. Weak utility usually means the token could disappear without changing much.
## How can beginners tell whether tokenomics are healthy or dangerous? [#how-can-beginners-tell-whether-tokenomics-are-healthy-or-dangerous]
Healthy tokenomics are usually simple to explain, transparent to verify, and aligned with long-term users rather than short-term insiders.
Here is a practical beginner checklist:
### Signs tokenomics may be healthier [#signs-tokenomics-may-be-healthier]
* A clear explanation of token purpose
* Reasonable circulating supply relative to total supply
* Transparent vesting schedule
* Gradual unlocks instead of giant cliffs
* Emissions tied to a real function like security or liquidity
* Supply changes that are easy to track
### Signs tokenomics may be dangerous [#signs-tokenomics-may-be-dangerous]
* Very low circulating supply with a much larger future supply
* Large insider allocations with short vesting
* Unclear or missing unlock calendar
* Utility that sounds vague or forced
* Yield that depends mostly on new token issuance
* Marketing focused on scarcity while ignoring dilution
If the tokenomics page feels designed to impress you instead of inform you, slow down.
That is also where broader risk research matters. Our guide on [how to spot a crypto rug pull](/blog/5-signs-crypto-rug-pull) pairs well with tokenomics because concentrated ownership and bad unlock design often show up together. And if you want more market context around supply narratives, our guides to [Bitcoin dominance](/blog/bitcoin-dominance-explained) and the [Fear & Greed Index](/blog/fear-greed-index-guide) help you separate fundamentals from mood.
## How does BlockMind help you research tokenomics faster? [#how-does-blockmind-help-you-research-tokenomics-faster]
BlockMind helps by turning tokenomics research into a conversation instead of a manual scavenger hunt. On Pro, BlockMind gives you a **personal AI investing agent** with its own workspace. You name it, connect the portfolio context you want it to use, and ask questions in plain English.
For example, ask: **"Explain the tokenomics of \[token] to me like I'm a beginner"** or "What could create sell pressure for this token over the next year?" Your agent digs into supply, unlocks, utility, and emissions, then answers in plain language. If you have connected a portfolio, it can explain the findings in the context of what you actually hold.
The practical difference is speed, clarity, and follow-through:
* Ask your agent to break down a token's supply, unlocks, utility, and likely sell pressure
* Use [Explore and Analyze](/docs/capabilities/research-journey) to find an idea and turn it into an analysis report
* Take that report to Verdicts, where a panel of experts returns a clear call: approved, watch, wait, or reject
* Track the assets you care about so material changes can appear in your Morning Brief on the dashboard and in your email inbox
For a quick check without an agent, run a free [DeepDive report](https://blockmind.app/deepdive). BlockMind uses wallet and exchange connections only to read balances and positions. Your agent cannot trade, withdraw, or move funds, and it will not tell you what to buy or sell. It does research, not financial advice; the decision stays with you.
That is the point of giving you an agent. Most crypto tools assume you already know how to interpret supply mechanics. Your BlockMind agent is built to explain them and keep watching after the first report.
## What should you do before buying a token? [#what-should-you-do-before-buying-a-token]
Before buying a token, check supply, unlocks, utility, and emissions in that order.
If you want a five-minute process, use this:
1. **Read the current circulating, total, and max supply**
2. **Find the vesting or unlock schedule**
3. **Identify what causes inflation, burns, or both**
4. **Explain the token's utility in one sentence**
5. **Ask what could create sell pressure in the next 3, 6, and 12 months**
6. **Check whether the story still makes sense after dilution**
If you cannot answer those questions, you do not understand the tokenomics yet.
And that is fine. It just means you should do more research before you take risk.
## Frequently Asked Questions [#frequently-asked-questions]
### What is crypto tokenomics explained simply? [#what-is-crypto-tokenomics-explained-simply]
Crypto tokenomics explained simply means understanding how a token's supply works, who receives tokens, when they unlock, and why the token has value inside a project. It is the supply-and-incentives layer behind a crypto asset.
### How do I read tokenomics as a beginner? [#how-do-i-read-tokenomics-as-a-beginner]
Read tokenomics as a beginner by checking four things first: circulating supply, max supply, vesting schedule, and token utility. If you understand those four, you understand most of the real risk.
### Is low circulating supply good in crypto? [#is-low-circulating-supply-good-in-crypto]
Low circulating supply is not automatically good. It can make a token look scarce today while hiding large future dilution from investor, team, or treasury unlocks.
### What is the difference between total supply and circulating supply? [#what-is-the-difference-between-total-supply-and-circulating-supply]
Circulating supply is the amount already tradable in the market. Total supply is the amount that currently exists overall, including locked tokens that are not circulating yet.
### Is Bitcoin inflationary or deflationary? [#is-bitcoin-inflationary-or-deflationary]
Bitcoin is best described as disinflationary. New BTC is still issued, but issuance falls over time through halvings, and the supply cap remains 21 million.
### Is Ethereum deflationary after EIP-1559? [#is-ethereum-deflationary-after-eip-1559]
Ethereum is not permanently deflationary in every condition. EIP-1559 burns the base fee, so ETH supply can become more deflationary during periods of high network activity, but net supply can still grow in other periods.
### How can I research tokenomics faster? [#how-can-i-research-tokenomics-faster]
You can research tokenomics faster by using a simple checklist or by handing the data gathering to a personal AI investing agent. Start with supply and unlocks, then require a plain-English analysis that can be checked against the underlying sources and monitored as the schedule changes.
## The Bottom Line [#the-bottom-line]
Good tokenomics research is not about memorizing vocabulary. It is about understanding who owns the supply, when more tokens can enter the market, what creates demand, and where future selling pressure may come from.
Run the same checks every time, and revisit them after you buy. Unlocks and emissions keep changing, so a token that looked healthy six months ago may carry a different risk today.
Use the broader [pre-buy crypto checklist](/blog/what-to-check-before-buying-crypto), then quantify the dilution gap with [market cap versus FDV](/blog/market-cap-vs-fdv-crypto) and inspect the actual [unlock and vesting schedule](/blog/token-unlocks-vesting-guide).
## Sources [#sources]
1. [Ethereum Improvement Proposal 1559](https://eips.ethereum.org/EIPS/eip-1559), April 2019.
2. [Bitcoin Developer Guide: Block Chain](https://developer.bitcoin.org/devguide/block_chain.html), accessed July 2026.
3. [Tokenomist: Token Unlocks and Vesting Data](https://tokenomist.ai/), accessed July 2026.
---
# How to Verify AI Crypto Analysis: The TRACE Protocol
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: problem-solving
Primary keyword: verify AI crypto analysis
Keywords: verify AI crypto analysis, AI crypto research accuracy, fact check AI crypto, AI hallucinations crypto research, crypto analysis verification
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Verify AI crypto analysis with TRACE: **Timestamp** the question and output, **Retrieve** the primary evidence, **Align** asset identity and metric definitions, **Calculate** every material number again, and **Expose** counterevidence, unknowns, and freshness limits. A confident answer with citations is still unverified until the cited evidence supports the exact claim.
***
## Why AI crypto analysis needs verification [#why-ai-crypto-analysis-needs-verification]
Crypto research combines volatile prices, changing supply, evolving contracts, pseudonymous addresses, provider-defined metrics, and issuer claims. Each creates a failure point even before AI is involved.
Generative AI adds another: it can produce fluent, confident statements that are false, internally inconsistent, or unsupported. NIST calls this **confabulation** and notes that it is a natural consequence of systems generating likely outputs rather than retrieving guaranteed facts ([NIST Generative AI Profile](https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf)).
Investor.gov warns not to rely solely on AI-generated information for investment decisions because inputs may be inaccurate, incomplete, outdated, or manipulated, and outputs can be faulty or made up ([Investor.gov](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/artificial-intelligence-fraud)).
The solution is not to avoid AI research. It is to give every material claim an evidence path that a human can reproduce.
This protocol complements [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research): your agent can gather, compare, monitor, and format; you verify sources, assumptions, incentives, and decisions.
**Scope and assumptions:** TRACE verifies research claims, not model internals or future market performance. It cannot make unavailable evidence public or prove private intent. It is not financial advice or a guarantee that a conclusion is correct.
*BlockMind’s agent will not tell you what to buy or sell and cannot trade, withdraw, approve, or move funds.*
## The TRACE verification protocol [#the-trace-verification-protocol]
### T — Timestamp the task, data, and output [#t--timestamp-the-task-data-and-output]
Record:
* Exact prompt or question
* Output time in UTC
* Asset, chain, and contract
* Requested research horizon
* Data timestamps for price, supply, flows, liquidity, and governance
* Source publication and retrieval dates
One timestamp is not enough. A current answer can cite an old whitepaper, a supply page revised last year, a price from minutes ago, and an audit for an obsolete deployment.
Use a freshness table:
| Claim type | Stale when… | Required timestamp |
| -------------------- | ------------------------------------------- | ---------------------------- |
| Price/depth | Market moved or quote expired | Observation time |
| Supply/unlocks | Contract, schedule, or provider changed | Data + retrieval time |
| Audit coverage | Deployment upgraded | Report + deployment revision |
| Team/governance | Roles or powers changed | Evidence + current check |
| Historical mechanism | Source superseded or implementation changed | Publication/version |
Do not use a universal expiry period. Define staleness by the event that changes the claim.
### R — Retrieve the primary evidence [#r--retrieve-the-primary-evidence]
For every material sentence, ask: **What is the closest source to the fact?**
| Claim | Preferred evidence |
| ------------------------- | ------------------------------------------------------------------ |
| Current contract behavior | Verified deployed source and decoded state |
| Token schedule | Vesting contract, executed governance, current official disclosure |
| Audit coverage | Auditor-hosted final report + exact revision |
| Partnership | Counterparty’s own announcement or filing |
| Market metric | Provider methodology + timestamped data |
| Transfer | Raw transaction + disclosed label method |
| Governance power | Current contracts, rules, and execution history |
| Repository activity | Canonical repository, pull requests, releases, deployment |
Open the source. Search for the exact field or passage. A citation can be real while supporting only part of the sentence.
Use four support statuses:
* **Direct:** Source explicitly supports the claim
* **Derived:** Claim follows from shown inputs and formula
* **Contextual:** Source is relevant but does not prove the claim
* **Unsupported:** Source is missing, inaccessible, stale, or mismatched
Only direct and reproducible derived evidence belong in a verified conclusion.
### A — Align identity, units, definitions, and scope [#a--align-identity-units-definitions-and-scope]
Before comparing or calculating, align:
* Correct token contract and chain
* Native versus wrapped or bridged asset
* USD versus token units
* UTC versus local dates
* Spot versus futures or perpetual markets
* Circulating versus unlocked, total, or maximum supply
* Gross versus net flow
* Address versus entity
* Revenue versus fees versus tokenholder income
* Audit commit versus live implementation
Create an alignment row:
```text
Claim C-07
Asset: [chain + contract]
Observation time: [UTC]
Unit: [token / USD / percentage]
Definition: [provider link]
Scope: [venues, contracts, wallets, period]
Exclusions: [explicit]
```
If two sources disagree, do not ask AI to “pick the accurate one” without tracing definitions. Preserve both values, methods, and likely reason for disagreement.
### C — Calculate material numbers independently [#c--calculate-material-numbers-independently]
Recompute:
* Market cap and FDV
* Percent changes and baselines
* Unlock as a share of current circulation
* Netflows
* Concentration shares
* Spreads and simulated impact
* Ratios comparing protocols
Show formula, inputs, units, and timestamp.
```text
Claimed unlock ratio = 8.4%
Formula = scheduled unlock / current circulating supply
Inputs = 42,000,000 / 500,000,000
Recalculation = 0.084 = 8.4%
Source dates = [schedule], [supply]
```
Then test whether the formula answers the claimed question. Correct arithmetic can still support a wrong inference. An unlock ratio does not calculate expected selling; exchange netflow does not calculate actual executed sales.
Use [Market Cap vs FDV](/blog/market-cap-vs-fdv-crypto), [Token Unlocks and Vesting](/blog/token-unlocks-vesting-guide), and [Crypto Liquidity Analysis](/blog/crypto-liquidity-analysis) for the relevant formulas and limitations.
### E — Expose counterevidence, uncertainty, and provenance limits [#e--expose-counterevidence-uncertainty-and-provenance-limits]
Require the answer to state:
* Strongest supporting evidence
* Strongest contradictory evidence
* Alternative explanations
* Unavailable evidence
* Invalidation condition
* Next freshness trigger
For a wallet alert:
> Observation: 10 million tokens moved to a provider-labeled exchange address.
> Alternative explanations: exchange internal transfer, custody, collateral, market-maker inventory, later withdrawal, or sale.
> Stronger evidence required for “sold”: decoded on-chain swap or off-chain execution evidence, plus entity-adjusted balances and market confirmation.
This is the discipline in [Whale Wallets and Exchange Flows](/blog/whale-wallets-exchange-flows).
## The claim verification ledger [#the-claim-verification-ledger]
| ID | AI claim | Material? | Source support | Definition aligned? | Math checked? | Counterevidence | Status |
| ---- | -------- | --------- | ------------------------------ | ------------------- | ------------- | --------------- | ------------------------------------------- |
| C-01 | | Yes/No | Direct/derived/contextual/none | Yes/No | Yes/No/N/A | | Verified/partial/unknown/contradicted/stale |
### Materiality test [#materiality-test]
A claim is material if changing it could change:
* The description of the project or token right
* A risk assessment
* A valuation or supply comparison
* A security conclusion
* The interpretation of market or on-chain behavior
* The next research action
Verify material claims first. Do not spend an hour checking a background date while the deployed-contract identity remains unknown.
## The source-quality ladder [#the-source-quality-ladder]
### Level 1: raw or executed primary evidence [#level-1-raw-or-executed-primary-evidence]
* Blockchain transaction and contract state
* Executed governance action
* Signed release or registry record
* Counterparty’s direct disclosure
### Level 2: current official documentation with scope [#level-2-current-official-documentation-with-scope]
* Protocol specification
* Tokenomics and vesting disclosure
* Auditor-hosted final report
* Data-provider methodology
### Level 3: independent high-authority analysis [#level-3-independent-high-authority-analysis]
* Regulator or standards body
* Reproducible research paper
* Transparent analytics provider
### Level 4: secondary reporting [#level-4-secondary-reporting]
* Reputable reporting that links evidence
* Analyst research with disclosed method
### Level 5: leads only [#level-5-leads-only]
* Social posts
* Screenshots
* Unattributed dashboards
* AI summaries without accessible sources
Higher is not always newer or complete. A current official supply page can be more useful than raw on-chain data if classification requires off-chain restrictions. Use the source closest to each claim and disclose its incentives.
## Worked hypothetical: verifying an AI token summary [#worked-hypothetical-verifying-an-ai-token-summary]
**“Pioneer Token” and all figures are fictional. This is not a real market analysis or recommendation.**
AI output:
> “Pioneer has a $300 million market cap, 90% of supply circulating, no major unlocks, deep liquidity, an active development team, and a completed audit. Whale withdrawals show accumulation.”
### T — Timestamp [#t--timestamp]
The answer has no data timestamp. Its supply citation is 14 months old; the audit is two years old.
### R — Retrieve [#r--retrieve]
* Current provider page reports 300 million circulating tokens at $1.
* Verified token contract and current docs show 1 billion maximum supply.
* A vesting contract makes 120 million tokens releasable in 45 days.
* Auditor report covers implementation version 1; live proxy uses version 3.
### A — Align [#a--align]
* The “90%” figure used total minted supply of 333 million as denominator, not 1 billion maximum supply.
* “Whale withdrawals” aggregate includes an exchange cold-wallet rotation.
* Development metric counts generated-data commits.
### C — Calculate [#c--calculate]
```text
Circulation ratio = 300m / 1bn = 30%, not 90%
Unlock / circulation = 120m / 300m = 40%
```
A $100,000 standardized sell quote shows 3.8% impact on the dominant venue, contradicting “deep liquidity” for that tested size.
### E — Expose [#e--expose]
Verified conclusion:
> Pioneer’s circulating market cap is $300 million at the timestamped $1 reference price under the cited provider definition. Circulation is 30% of stated maximum supply. A tranche equal to 40% of current circulation becomes releasable in 45 days, although sale intent is unknown. Current audit coverage and the accumulation interpretation are unverified; tested liquidity is constrained at $100,000.
TRACE changed nearly every strong adjective without predicting price.
## Citation verification checklist [#citation-verification-checklist]
For each source:
* Does the URL resolve to the publisher, not a copied page?
* Is the document final and current?
* Does it refer to the same chain, contract, version, and date?
* Does the cited passage support the entire sentence?
* Is the claim a quote, paraphrase, or calculation?
* Are qualifiers and exclusions preserved?
* Does another primary source contradict it?
* Is an archive or transaction hash needed for reproducibility?
Do not overquote. Preserve enough context to verify without reproducing protected material.
## Visual and media claims [#visual-and-media-claims]
Screenshots and videos can be stale, edited, impersonated, or detached from context. Verify through the publisher’s current official channel and the underlying source.
C2PA develops provenance standards that can help validate assertions about a media asset’s origin and editing history. Its principles also state that validated provenance does not decide whether the content is “good” or “bad” ([C2PA](https://c2pa.org/principles/)). Provenance can support authenticity; it does not prove the claim inside the media.
For a partnership screenshot, verify the counterparty. For a dashboard screenshot, open the live dashboard and methodology. For an explorer image, retrieve the transaction hash.
## How to ask an AI agent for verifiable research [#how-to-ask-an-ai-agent-for-verifiable-research]
Use a prompt that requires evidence structure:
> “Research \[asset, chain, contract] as of \[UTC time]. Separate observations from interpretations. Give each material claim an ID, link the closest primary source, quote or identify the exact supporting field, state data time and definition, show calculations, include the strongest counterevidence and alternative explanations, and mark inaccessible evidence unknown. Do not infer wallet intent or audit coverage.”
Then ask for an adversarial pass:
> “Try to disprove claims C-01 through C-05 using current primary sources. Report definition mismatches, stale evidence, scope gaps, and calculations that answer a different question.”
Use the copyable [Crypto Research Workflow Template](/blog/crypto-research-workflow-template) to preserve the output.
## Red flags in AI-generated crypto research [#red-flags-in-ai-generated-crypto-research]
* No asset contract or chain
* “Current” numbers without timestamps
* Citations collected at the end but not tied to claims
* A source that repeats the project’s press release presented as independent confirmation
* Percentages without denominators
* Market cap described as money invested
* FDV described as future value
* Unlocks described as guaranteed selling
* Exchange transfers described as buys or sales
* Audit badge described as security
* Commit count described as adoption
* Price predictions or guaranteed-return language
* Confidence percentages with no calibrated methodology
* No counterevidence or unknowns
The CFTC specifically warns that AI cannot predict future or sudden market changes and cautions against systems promising guaranteed or unreasonable returns ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)). Treat certainty marketing as a safety warning, not evidence.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Primary sources can be wrong, self-interested, or corrected later.
* On-chain evidence does not reveal all beneficial ownership or intent.
* Data-provider methodologies can be proprietary.
* A verified calculation can still be economically irrelevant.
* Adversarial checking can miss shared assumptions.
* Some claims cannot be verified publicly.
* TRACE improves claim reliability; it cannot predict markets or eliminate risk.
Mark unresolved claims as unresolved. That is a valid research outcome.
## How BlockMind supports verification [#how-blockmind-supports-verification]
A BlockMind agent can retrieve current sources, inspect live pages through its [browser capability](/docs/capabilities/browser), review public wallet and protocol data with [on-chain intelligence](/docs/capabilities/on-chain), and save the TRACE ledger to your Notebook. Its work remains visible as research material rather than an instruction to act.
Ask your agent to attach source links, timestamps, definitions, calculations, and counterevidence. Then manually open the few sources that carry the conclusion.
## The Bottom Line [#the-bottom-line]
Fluency is not verification. Use TRACE: timestamp every volatile input, retrieve the closest primary evidence, align identity and definitions, recalculate material numbers, and expose contradictions and unknowns.
The goal is not to prove the AI right. It is to produce a narrower, reproducible research record that remains trustworthy even when the original answer was wrong.
## Sources [#sources]
* [NIST — Generative AI Profile, Including Confabulation Risk](https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf)
* [NIST — AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework)
* [Investor.gov — Artificial Intelligence and Investment Fraud](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/artificial-intelligence-fraud)
* [CFTC — AI Won’t Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
* [C2PA — Guiding Principles for Content Provenance](https://c2pa.org/principles/)
* [C2PA — Specifications](https://spec.c2pa.org/about/)
---
# Liquidity Locks in Crypto Explained: What They Protect—and What They Don't
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: informational
Primary keyword: liquidity locks crypto
Keywords: liquidity locks crypto, locked liquidity meaning, how to check liquidity lock, LP token lock, crypto liquidity lock explained
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A liquidity lock prevents the owner of a specific liquidity position from withdrawing that position until a stated time. It does not lock every pool, prevent token minting or blacklisting, vest team tokens, guarantee adequate exit depth, or prove the project is legitimate. Verify the pool, position, owner, percentage, and unlock timestamp on-chain.
***
“Liquidity locked” is one of the most reassuring—and most incomplete—claims on a token launch page.
A real lock can remove one common rug mechanism: the main liquidity provider redeeming the position and draining both assets from the pool. But the protection is only as broad as the position placed in the locker. If half the liquidity is locked, the other half can still move. If the lock covers an inactive pool, it may protect nothing useful. If the token contract can block sales or mint unlimited supply, locked liquidity does not remove those risks.
## What is a crypto liquidity lock? [#what-is-a-crypto-liquidity-lock]
A crypto liquidity lock places the ownership token for a decentralized-exchange liquidity position into a time-locked smart contract. Until the unlock timestamp, the locker prevents the lock owner from retrieving that position under the lock's documented rules.
[Team Finance's current liquidity-lock explainer](https://blog.team.finance/how-liquidity-locks-prevent-rug-pulls-and-how-strong-projects-use-them-correctly/) describes the process for fungible LP tokens: the holder sends them to a time-locked smart contract and can claim them after expiry. [UNCX documentation](https://docs.uncx.network/archives/liquidity-lockers-v2-legacy-app) similarly describes a locker as restricting the developer's ability to move LP tokens between a start and end date.
The lock is not the liquidity pool itself. It controls the asset that represents ownership of some portion of that pool.
## LP tokens, NFTs, and why DEX version matters [#lp-tokens-nfts-and-why-dex-version-matters]
On [Uniswap v2](https://developers.uniswap.org/docs/get-started/concepts/how-uniswap-works), liquidity providers receive fungible ERC-20 pool tokens representing a proportional reserve share. Uniswap v3 and v4 positions are represented by NFTs because each position can have its own price range and parameters.
Uniswap's support documentation states that the wallet holding the UNI-V2 tokens or position NFT owns the liquidity and can remove it, claim fees, or add to the position. Transferring the ownership token transfers control.
That changes verification:
* **v2-style pool:** identify total LP-token supply, the amount in the locker, and other holders.
* **v3/v4-style pool:** identify the exact position NFT, its range, liquidity value, owner, and whether other meaningful positions remain unlocked.
A website badge saying “LP locked” does not tell you which version, which pool, or which position.
## What a liquidity lock protects [#what-a-liquidity-lock-protects]
When correctly configured, a lock can prevent its owner from:
* redeeming the locked LP tokens or NFT before expiry;
* transferring the locked position during the lock period, subject to the locker design;
* performing the classic immediate liquidity-removal rug with that position.
This is useful evidence about one control at one time. Chainalysis' market-manipulation methodology included large liquidity removal followed by an inactive pool as a pattern for suspected pump-and-dump or rug activity. Making the main position unavailable for removal can reduce that specific opportunity.
## What a liquidity lock does not protect [#what-a-liquidity-lock-does-not-protect]
A lock does **not** establish that:
* all liquidity is locked;
* the lock covers the active or deepest pool;
* the lock lasts long enough for your time horizon;
* a v3 position covers the current price range;
* another unlocked provider cannot remove liquidity;
* the token can be sold;
* taxes, blacklists, maximum-sell limits, or trading status cannot change;
* supply cannot be minted;
* insiders' token allocations are vested;
* an admin or proxy implementation cannot change;
* the locker contract itself has no defect;
* the team will deliver a product;
* the market price is not manipulated.
Liquidity risk and contract risk are separate columns in a due-diligence table.
## The LOCK verification framework [#the-lock-verification-framework]
### L — Locate the live market [#l--locate-the-live-market]
Start with the token's exact contract address. Find every meaningful trading pair and record:
* DEX and version;
* pool address;
* paired asset;
* current liquidity;
* recent volume;
* price range for concentrated positions;
* share of actual trading occurring there.
Do not let the project choose the pool you inspect. The economically relevant pool is the one traders actually rely on.
### O — Ownership and locker contract [#o--ownership-and-locker-contract]
Identify who owns each liquidity position. Then confirm that the apparent locker address belongs to the claimed service using the locker's official app or documentation.
Record:
* locker contract address;
* lock transaction hash;
* lock owner or beneficiary;
* LP token amount or NFT token ID;
* ability to split, extend, transfer ownership, or partially claim;
* whether the locker supports that DEX version.
Team Finance documents features such as split locks, ownership transfers, and partial claims. These features may be legitimate administration, but they mean the lock record must be read precisely rather than reduced to a badge.
### C — Coverage and calendar [#c--coverage-and-calendar]
For a fungible LP-token pool, estimate lock coverage as:
`locked LP tokens ÷ total outstanding LP tokens × 100`.
**Hypothetical example:** A pool has 10,000 LP tokens outstanding. The verified locker holds 7,500, the deployer holds 1,500, and other providers hold 1,000. Lock coverage is `7,500 ÷ 10,000 × 100 = 75%`. The claim “liquidity locked” is technically true, but 25% remains outside the lock and the deployer directly controls 15%.
Coverage is not identical to dollar value if positions or versions differ. For concentrated liquidity, calculate coverage using the active liquidity around the current price, not merely NFT count.
Also record the exact UTC unlock time and calculate `days remaining = unlock timestamp - verification timestamp`. A lock that expires tomorrow and a lock that lasts years should not receive the same label.
### K — Known escape routes [#k--known-escape-routes]
Search for ways the visible protection could be bypassed:
* new pool with unlocked liquidity;
* migration to another token or contract;
* proxy upgrade;
* minting and dumping new tokens into the locked pool;
* transfer restrictions or extreme sell tax;
* unlocked team supply large relative to pool depth;
* administrative function that changes pairs or exemptions;
* lock expiry or partial claim;
* concentrated position moving out of range.
The LOCK framework ends with adversarial thinking because a real lock can coexist with a different rug path.
## Lock coverage is not exit coverage [#lock-coverage-is-not-exit-coverage]
Even permanently locked liquidity can be too shallow for holders to exit without severe price impact. Measure available depth against the position, not just whether LP ownership is time-locked.
Use a simple stress ratio:
`position value you may need to exit ÷ quoted pool depth within your acceptable price impact`.
**Hypothetical:** A wallet holds tokens marked at $20,000. The pool has only $5,000 of executable depth within a 10% price-impact band. The stress ratio is `20,000 ÷ 5,000 = 4`. The displayed value is four times the quoted depth within that band. A 100% lock does not create the missing buyer-side liquidity.
This is not a universal risk threshold. It is a reminder that nominal portfolio value and executable exit value differ.
## Liquidity lock vs token vesting [#liquidity-lock-vs-token-vesting]
These controls protect different assets:
| Control | Asset restricted | Primary risk addressed |
| ------------------- | ------------------------------ | -------------------------------------------- |
| Liquidity lock | LP tokens or position NFT | Withdrawal of that pool liquidity |
| Team-token lock | The project's token allocation | Immediate insider token sales |
| Vesting schedule | Tokens released over time | Supply cliffs and incentive alignment |
| Timelock controller | Administrative actions | Surprise parameter or implementation changes |
A project can lock 100% of its LP position while team wallets hold a large, liquid token allocation. It can also vest team tokens while leaving pool liquidity removable. Check both.
## Locked vs burned liquidity [#locked-vs-burned-liquidity]
“Burned liquidity” usually means sending fungible LP tokens to an address from which they are assumed unrecoverable. That can make the position effectively permanent. It also removes the ability to recover or migrate it through ordinary ownership.
Do not assume every apparent burn address is inaccessible or that every DEX version uses fungible LP tokens. Verify the address, token standard, chain, and position mechanics. A v3 position NFT requires different analysis from v2 LP tokens.
A time lock is reversible after expiry. A genuine burn is intended to be irreversible. Neither protects against malicious token logic.
## A due-diligence table for any lock [#a-due-diligence-table-for-any-lock]
| Field | Evidence required | Red flag |
| --------------- | ------------------------------------------------- | ------------------------------------ |
| Token | Exact contract/mint address | Search by ticker only |
| Pool | On-chain pair address and DEX version | Unclear or inactive pool |
| Position | LP-token amount or NFT ID and range | Badge without position data |
| Owner | Locker contract verified through official service | Project-controlled wallet |
| Coverage | Share of outstanding/active liquidity | Small fraction presented as “locked” |
| Duration | Exact unlock timestamp | Short or already expired lock |
| Other owners | Full LP holder/position map | Large deployer-controlled remainder |
| Other pools | Liquidity and volume across venues | Deeper unlocked pool elsewhere |
| Contract powers | Mint, tax, blacklist, pause, proxy roles | Unexplained mutable controls |
Save the table with block number and timestamp. Liquidity ownership changes.
## Common misleading claims [#common-misleading-claims]
### “100% liquidity locked” [#100-liquidity-locked]
Ask: 100% of whose liquidity? It may mean 100% of the team's position, not 100% of pool liquidity. It may also refer to one pool while another is active.
### “Locked forever” [#locked-forever]
Verify the actual timestamp or burn transaction. Marketing language is not an on-chain state.
### “Audited locker” [#audited-locker]
An audit of the locker does not audit the token, pool economics, team wallets, or integration. Match report scope and deployed addresses.
### “Liquidity cannot be rugged” [#liquidity-cannot-be-rugged]
Perhaps the locked position cannot be removed before expiry. The token can still be made unsellable, diluted, dumped, abandoned, or migrated.
## Limitations and counterevidence [#limitations-and-counterevidence]
Not every unlocked pool is a scam. Professional liquidity management, market making, migrations, and concentrated-liquidity repositioning can require flexibility. Conversely, a long lock can be used as persuasive theater around a malicious token.
The correct conclusion is narrow: a verified lock restricts a specified owner and position according to a specified contract until a specified time. Whether the overall project is investable requires the rest of the [rug pull checker workflow](/blog/crypto-rug-pull-checker-guide) and [manual DYOR checklist](/blog/manual-dyor-checklist-12-steps).
Also verify that the locked pool remains practically sellable using the [honeypot token workflow](/blog/honeypot-crypto-token); a liquidity lock cannot override restrictive transfer logic.
## How BlockMind can help [#how-blockmind-can-help]
A BlockMind agent can help collect pool addresses, locker records, contract powers, holder concentration, and contradictory claims into a repeatable analysis. It can store the evidence in your Notebook and monitor public market context. It cannot prove a locker's code is flawless, identify every linked wallet, or certify a token. BlockMind never trades or moves funds.
## The Bottom Line [#the-bottom-line]
Liquidity locks matter, but only within their exact scope. Verify **what** is locked, **where**, **how much**, **by whom**, and **until when**. Then investigate the routes the lock does not cover.
If a project offers a screenshot instead of a pool address, position ID, locker contract, and timestamp, the claim is not yet evidence.
This article is for research and education, not financial advice.
## Sources [#sources]
1. [Uniswap: How liquidity positions are represented](https://support.uniswap.org/hc/en-us/articles/20980786685069-Why-is-liquidity-position-ownership-represented-by-tokens-or-NFTs)
2. [Uniswap developers: How Uniswap works across v2, v3, and v4](https://developers.uniswap.org/docs/get-started/concepts/how-uniswap-works)
3. [Team Finance: How Liquidity Locks Prevent Rug Pulls](https://blog.team.finance/how-liquidity-locks-prevent-rug-pulls-and-how-strong-projects-use-them-correctly/), March 2026.
4. [UNCX: Liquidity-locker documentation](https://docs.uncx.network/archives/liquidity-lockers-v2-legacy-app)
5. [OpenZeppelin: Access control and timelocks](https://docs.openzeppelin.com/contracts/5.x/access-control)
6. [Chainalysis: Pump-and-dump and liquidity-removal methodology](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/)
---
# Manual DYOR Checklist: The 12 Steps Most People Skip
Published: 2026-03-10
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: crypto DYOR checklist
Keywords: DYOR checklist, crypto research checklist, crypto due diligence
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A real DYOR checklist means checking the team, whitepaper, tokenomics, audits, holders, liquidity, sentiment, on-chain activity, competition, community, technicals, and risk/reward before you buy. Most people skip half of that work because it takes too long, which is exactly why bad tokens keep attracting capital.
***
If you want to know **how to DYOR**, start here: proper crypto due diligence is not one check, one chart, or one influencer thread. It is a stack of verification steps that reduce the chance of buying into hype, bad token design, manipulation, or outright fraud.
That matters because the cost of bad research is still high. As of 2025, [Chainalysis estimated](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/) that suspected wash trading on select blockchains may account for up to **$2.57 billion in trading volume**, which is a useful reminder that volume and hype can be manufactured rather than earned. In practice, the harder a token is to verify across multiple angles, the more careful you should be.
This guide breaks down a **manual DYOR checklist crypto investors can actually use**. It also shows why serious research becomes time-consuming fast, especially if you are checking multiple tokens or trying to understand how a new position fits your existing portfolio.
If you want a broader framework for where AI fits into that process, see our pillar on [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research).
## What is a proper DYOR checklist in crypto? [#what-is-a-proper-dyor-checklist-in-crypto]
A proper DYOR checklist in crypto is a repeatable process for verifying whether a token is credible, how risky it is, and what would need to go right for the investment to work.
Most retail investors reduce DYOR to reading a website, checking the chart, and scrolling X for five minutes. That is not research. Real due diligence means cross-checking the project from multiple angles: the people behind it, the token structure, the smart contracts, the holder base, the liquidity, the actual on-chain usage, and the quality of the community around it.
The goal is not to find a “perfect” project. The goal is to spot hidden risk before your capital is exposed.
## 1) How do you verify the team behind a crypto project? [#1-how-do-you-verify-the-team-behind-a-crypto-project]
You verify the team by checking whether the founders, builders, and advisors are real people with a traceable history of shipping relevant work.
Start with the obvious checks: LinkedIn, X, GitHub, previous projects, podcasts, interviews, conference appearances, and legal entity information. Then go one step deeper. Do the founders have a consistent history across platforms? Do past teammates mention them? Did they actually build the products they claim to have built?
**What to look for:**
* Real identities, not only cartoon avatars
* Verifiable work history in crypto, finance, or software
* GitHub activity or public technical contributions
* Consistent bios across LinkedIn, X, company site, and media mentions
* Advisors who are actually acknowledged by the advisors themselves
**Red flags:**
* Fake or empty LinkedIn profiles
* Stolen profile photos
* No evidence of previous execution
* Team pages full of vague titles and zero specifics
* “Stealth” used as an excuse to avoid accountability
Anonymous teams are not automatically scams. But anonymous teams remove one layer of accountability, so the rest of your checklist needs to be much stronger.
## 2) How should you analyze a crypto whitepaper? [#2-how-should-you-analyze-a-crypto-whitepaper]
You should analyze a crypto whitepaper by asking whether it explains a real problem, a believable solution, and a working mechanism instead of just selling a narrative.
A strong whitepaper makes the business model, token role, and system design easier to understand. A weak whitepaper hides behind buzzwords like “revolutionary,” “community-driven,” or “AI-powered” without explaining how value is created or captured.
**What to look for:**
* Clear explanation of the problem being solved
* Specific description of how the protocol or product works
* Real token utility, not generic “governance” filler
* Transparent roadmap with technical milestones
* Risks, assumptions, and tradeoffs explained honestly
**Red flags:**
* No whitepaper at all
* Whitepaper reads like marketing copy
* Heavy jargon with no mechanism behind it
* Copy-paste sections from other projects
* Claims that the token will rise because the community is strong
If you cannot explain the project in plain language after reading the whitepaper, that is usually a sign the project either lacks clarity or is hiding complexity behind narrative.
## 3) What should you check in tokenomics? [#3-what-should-you-check-in-tokenomics]
You should check who gets the tokens, when they unlock, what the token is used for, and whether supply mechanics create sustainable demand or constant sell pressure.
Tokenomics is where a lot of otherwise interesting projects break. A project can have a compelling story and still be a bad investment if insiders own too much, unlocks are aggressive, or the token has weak utility.
**What to look for:**
* Circulating supply versus fully diluted valuation
* Team, investor, treasury, and community allocations
* Vesting schedules and unlock timelines
* Real token demand drivers such as fees, collateral, or usage
* Emissions, burns, and inflation dynamics
**Red flags:**
* Tiny circulating supply paired with huge FDV
* Large insider allocations with short vesting
* Token exists mainly for fundraising, not product use
* Yield or rewards paid by constant token inflation
* No credible plan for post-incentive demand
This is one of the easiest places to get fooled. A token can look “cheap” on price alone while being structurally expensive once future supply hits the market.
## 4) How do you check whether the smart contract is safe? [#4-how-do-you-check-whether-the-smart-contract-is-safe]
You check whether the contract has been audited, whether the audit is credible, and whether the deployed contract actually matches what was reviewed.
An audit is not a guarantee of safety, but it is still a useful filter. The right question is not “Does the project have an audit?” The right question is “What did the audit cover, who performed it, and were the issues fixed?”
**What to look for:**
* Public audit report from a known security firm
* Audit date and contract scope clearly stated
* Evidence that critical issues were fixed
* Verified contract source code on a block explorer
* Admin privileges, pause functions, mint rights, and upgradeability explained
**Red flags:**
* “Audited” with no report link
* Audit from an unknown firm with no reputation
* Audit covers old code, not deployed code
* Owner functions can change fees, block wallets, or mint freely
* Contract is unverified on-chain
For high-risk tokens, this step alone is not enough. You also need to understand what powers the team still controls after launch.
## 5) Why does holder distribution matter? [#5-why-does-holder-distribution-matter]
Holder distribution matters because concentrated ownership creates concentration risk, and concentration risk can crush price even when the story still sounds good.
If a handful of wallets control a large share of supply, those holders can move the market, dump into rallies, or coordinate exits before retail reacts.
**What to look for:**
* Percentage held by top 10 and top 50 wallets
* Whether top wallets are exchanges, treasury, burned supply, or insiders
* Vesting wallets versus liquid wallets
* Changes in top-holder concentration over time
* Whale accumulation or distribution patterns
**Red flags:**
* A few non-exchange wallets hold a massive share of supply
* Team wallets are not disclosed
* Treasury holdings are presented as “community owned” without governance transparency
* Repeated transfers between related wallets create a false impression of decentralization
If you want a shorter warning-sign version of this step, our guide on [5 signs a crypto project might be a rug pull](/blog/5-signs-crypto-rug-pull) covers the basics.
## 6) How do you assess token liquidity? [#6-how-do-you-assess-token-liquidity]
You assess liquidity by checking whether you can realistically enter and exit a position without massive slippage or team-controlled liquidity risk.
A token can show attractive volume but still be dangerous if most of that activity is shallow, fragmented, or manipulated. Liquidity is what determines whether the market is actually tradeable.
**What to look for:**
* Liquidity depth in major trading pairs
* Where liquidity sits: DEXs, CEXs, or both
* Slippage on realistic order sizes
* Locked versus removable LP liquidity
* Concentration of market-making across a small number of wallets or venues
**Red flags:**
* Small liquidity pools relative to market cap
* Huge reported volume with poor depth
* Liquidity mostly controlled by the team
* LP tokens not locked or ownership unclear
* Wide spread and sharp price impact on modest trades
This is where manual research gets tedious fast. You often need to compare multiple pools, venues, and wallet movements before you can trust the picture.
## 7) What should you check in social sentiment? [#7-what-should-you-check-in-social-sentiment]
You should check whether attention is organic, whether the conversation is improving or deteriorating, and whether the audience sounds informed or purely promotional.
Social sentiment matters because early hype can attract liquidity, but low-quality hype also attracts bad actors. The signal is not “Are people talking about it?” The signal is “What kind of people are talking about it, and what exactly are they saying?”
**What to look for:**
* Growth in mentions across X, Telegram, Discord, and Reddit
* Whether discussion is technical, product-focused, and specific
* Developer updates versus only marketing updates
* Balance between bullish sentiment and critical discussion
* Whether community questions get answered clearly
**Red flags:**
* Engagement dominated by giveaways and price targets
* Comment sections full of bots, copy-paste replies, or generic hype
* Community cannot explain the product beyond meme slogans
* Founders post constantly about price and rarely about shipping
* Sudden attention spike with no product or usage catalyst
Sentiment is useful, but it should confirm your thesis, not create it.
## 8) How do you review on-chain activity? [#8-how-do-you-review-on-chain-activity]
You review on-chain activity by checking whether real users, capital, and transactions support the story the project tells.
This is where narrative meets evidence. If a protocol claims traction, there should be some visible combination of active wallets, transaction count, fee generation, TVL, retention, or contract interactions to support that claim.
**What to look for:**
* Growth or decline in active users and transactions
* TVL or usage metrics that match the product type
* Fee generation or revenue where relevant
* Repeat activity rather than one-off bursts
* Whale behavior and smart-money wallet flows
**Red flags:**
* Usage flat or falling while social hype is rising
* TVL inflated by mercenary incentives
* Wash-like activity that creates fake engagement
* Large wallet inflows with no genuine user growth
* No sign that the token is connected to the actual product
This is where an AI analyst can beat spreadsheets. With a Pro trial or subscription, a BlockMind agent pulls a token's fundamentals, on-chain signals, risk, and sentiment into one analysis report, then keeps watching the idea after you save it. For a quick one-off check without an agent, a free [DeepDive report](https://blockmind.app/deepdive) gives you a structured first pass on any coin in minutes.
## 9) How do you analyze the competitive landscape? [#9-how-do-you-analyze-the-competitive-landscape]
You analyze the competitive landscape by asking whether the project is meaningfully better, faster, cheaper, or more defensible than the alternatives already live.
Many tokens sound impressive until you compare them with direct competitors. A project may not need to be number one, but it does need a believable reason to win part of the market.
**What to look for:**
* Direct competitors in the same category
* Clear wedge such as distribution, product quality, speed, or economics
* Network effects, switching costs, or ecosystem advantages
* Market size relative to current valuation
* Whether the token captures value if the product succeeds
**Red flags:**
* “No competitors” as a positioning claim
* Product is basically a clone with weaker traction
* Token value depends on category growth alone
* The bull case assumes perfect execution in a crowded field
* No moat except community belief
A good project can still be a bad bet if the category is saturated and the token capture is weak.
## 10) What makes a crypto community high quality? [#10-what-makes-a-crypto-community-high-quality]
A high-quality crypto community is one that improves your understanding of the project instead of only increasing your excitement about the price.
The best communities create feedback loops between users, builders, and researchers. They surface bugs, explain changes, challenge bad assumptions, and attract long-term participants.
**What to look for:**
* Users discussing product updates and real use cases
* Mods and founders answering tough questions directly
* Community members creating thoughtful analysis or tutorials
* Signs of retention, not only campaign-driven bursts
* Civil disagreement without censorship of every critical point
**Red flags:**
* Every conversation routes back to “wen moon”
* Legitimate questions are dismissed as FUD
* Moderation exists mainly to protect price narrative
* Community engagement collapses when price cools off
* No sign that actual users are present
Community quality is underrated because it is hard to quantify. But it often tells you whether a project is building conviction or just renting attention.
## 11) Why should technical indicators be part of DYOR? [#11-why-should-technical-indicators-be-part-of-dyor]
Technical indicators should be part of DYOR because even strong projects can be bad entries if momentum, trend structure, or market regime are against you.
This step does not replace fundamentals. It helps with timing and context. If the token is fundamentally interesting but breaking key support on weak volume while the broader market is risk-off, the better decision may be to wait.
**What to look for:**
* Trend direction on higher timeframes
* Support and resistance zones
* Relative strength versus BTC and sector peers
* Volume confirmation on breakouts and breakdowns
* Broader market context such as Fear & Greed, Altcoin Season, and BTC Dominance
**Red flags:**
* Buying after a vertical move with no consolidation
* Ignoring market regime entirely
* Taking low-liquidity chart patterns too seriously
* Entering because “it already ran, so it must keep running”
If you need help understanding regime context, our guides on [Fear & Greed](/blog/fear-greed-index-guide) and [Bitcoin dominance](/blog/bitcoin-dominance-explained) are useful starting points. BlockMind refreshes those readings once daily on its free Market Indicators page. With a Pro agent, the same readings appear in the dashboard's indicator carousel and can be folded into your Morning Brief as market context. The [market indicators docs](/docs/classic/market-indicators) explain how each one works.
## 12) How do you calculate risk/reward before buying? [#12-how-do-you-calculate-riskreward-before-buying]
You calculate risk/reward by defining what could go right, what could go wrong, what would invalidate your thesis, and how much downside you can actually tolerate.
This is the step most people skip because it forces honesty. It turns “I like the project” into a measurable decision.
**What to look for:**
* Base case, bull case, and bear case
* Clear thesis for why the token should outperform
* Specific invalidation points
* Position size relative to your portfolio
* Liquidity-adjusted exit plan
**A simple framework:**
* What needs to happen for this token to work?
* What would prove me wrong?
* How much can I lose if I am early or simply wrong?
* Is the upside large enough compared with that downside?
* Does this position improve my portfolio, or just add more of the same risk?
A token can pass several checklist steps and still fail the risk/reward test. That is not a contradiction. That is discipline.
## Which of These 12 Steps Can an AI Agent Run for You? [#which-of-these-12-steps-can-an-ai-agent-run-for-you]
After finishing the checklist, turn your notes into a decision: buy, wait, avoid, or re-check later when new data arrives.
That sounds simple, but it is where most manual research breaks down. By the time you finish checking all 12 steps, the market may have moved, your conviction may still be fuzzy, and you may have three more tokens to analyze.
That is the real bottleneck. Manual DYOR is not hard because any one step is impossible. It is hard because doing **all of them consistently** takes time, attention, and the ability to connect scattered signals into one view.
The useful split is to let an agent gather and organize evidence while you keep the decisions that require judgment:
| Checklist Area | What an Agent Can Do | What Stays With You |
| ------------------------------------------------ | ---------------------------------------------------------- | ---------------------------------------------------------- |
| Team and whitepaper (steps 1–2) | Gather backgrounds, claims, and source material | Judge credibility, clarity, and founder quality |
| Tokenomics through on-chain activity (steps 3–8) | Compare supply, holders, liquidity, sentiment, and usage | Verify audits and liquidity locks at their primary sources |
| Competition and community (steps 9–10) | Summarize alternatives and patterns in discussion | Judge the moat, incentives, and community quality |
| Technical and market context (step 11) | Refresh indicators and connect them to the asset | Decide whether the regime changes your thesis |
| Risk and reward (step 12) | Surface concentration, correlation, and invalidation risks | Set position size and make the final decision |
With a Pro trial or subscription, BlockMind gives you a **personal AI investing agent** with its own workspace. Your agent can produce analysis reports, send saved reports to a committee of expert frameworks for a Verdict (approved, watch, wait, or reject), connect research to your [portfolio](/docs/data/portfolios) and [watchlists](/docs/data/watchlists), and send a Morning Brief to your dashboard and email every day.
BlockMind uses wallet and exchange connections only to read balances and positions. The agent can't trade, withdraw, or move funds even if you told it to. It does research, not financial advice, and the final decision stays with you.
If you want a one-off report first, run a free DeepDive report on any token.
## Frequently Asked Questions [#frequently-asked-questions]
### What is DYOR in crypto? [#what-is-dyor-in-crypto]
DYOR means “Do Your Own Research.” In crypto, it refers to independently verifying whether a token or project is credible, risky, overhyped, or worth deeper consideration before you invest.
### What is the best DYOR checklist crypto investors should follow? [#what-is-the-best-dyor-checklist-crypto-investors-should-follow]
The best DYOR checklist crypto investors should follow covers 12 areas: team verification, whitepaper analysis, tokenomics, audit checks, holder distribution, liquidity, social sentiment, on-chain activity, competition, community quality, technical indicators, and risk/reward.
### How to DYOR before buying a crypto token? [#how-to-dyor-before-buying-a-crypto-token]
To DYOR before buying a crypto token, verify who built it, read the whitepaper critically, inspect tokenomics, check audits, review holders and liquidity, study on-chain activity and sentiment, compare competitors, and define a clear risk/reward plan before entering.
### How long does proper crypto DYOR take? [#how-long-does-proper-crypto-dyor-take]
Proper crypto DYOR can take anywhere from 30 minutes for a rough filter to several hours for a serious investment case. The deeper the position size and the riskier the token, the more time you should expect to spend.
### Can AI replace manual DYOR? [#can-ai-replace-manual-dyor]
AI should not replace judgment, but it can dramatically reduce the manual work involved in gathering, organizing, and explaining the signals that matter. That is the model behind BlockMind's personal agent: it gathers and structures the evidence in this checklist while the final decision stays with you.
### Is a whitepaper enough for DYOR? [#is-a-whitepaper-enough-for-dyor]
No. A whitepaper is only one input. It tells you the story the project wants to tell. DYOR starts when you test that story against tokenomics, contracts, holders, liquidity, and actual on-chain behavior.
### What are the biggest mistakes people make when doing DYOR? [#what-are-the-biggest-mistakes-people-make-when-doing-dyor]
The biggest mistakes are trusting hype, ignoring unlocks, skipping holder concentration, assuming volume equals liquidity, treating audits as guarantees, and buying without a defined invalidation point.
## The Bottom Line [#the-bottom-line]
A DYOR checklist only protects you if you complete it consistently. Use AI to compress the gathering, cross-referencing, and monitoring, but verify security claims at their primary sources and keep every sizing decision for yourself.
Start with the shorter [pre-buy evidence checklist](/blog/what-to-check-before-buying-crypto), use the [whitepaper verification process](/blog/how-to-read-crypto-whitepaper), and record the work in the [crypto research workflow template](/blog/crypto-research-workflow-template).
## Sources [#sources]
1. [Chainalysis: Market Manipulation, Wash Trading, and Pump-and-Dump Schemes](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/), 2025.
2. [SEC Investor.gov: Crypto Asset Scam Warning Signs](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/crypto-scams), May 2024.
3. [CFTC: Customer Advisory on Virtual Currency Pump-and-Dump Schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html), accessed July 2026.
---
# Crypto Market Cap vs FDV: Formulas, Traps, and a Better Comparison
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: comparison
Primary keyword: market cap vs FDV
Keywords: market cap vs FDV crypto, fully diluted valuation crypto, crypto market capitalization formula, FDV market cap ratio, crypto dilution
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Crypto market cap estimates the value of the currently circulating supply at the current reference price. Fully diluted valuation estimates the value of a larger supply—usually maximum supply—at that same price. Neither is a forecast, neither measures cash invested, and neither is enough alone. The useful analysis adds the unlock timeline, recipient incentives, liquidity, supply-definition confidence, and demand needed to absorb new circulation.
***
## What is the difference between crypto market cap and FDV? [#what-is-the-difference-between-crypto-market-cap-and-fdv]
The common formulas are:
```text
Circulating market cap = current reference price × circulating supply
FDV = current reference price × maximum supply
```
Some providers use total supply when a meaningful maximum supply is unavailable, so always read the methodology. CoinMarketCap’s published method distinguishes circulating market cap, unlocked market cap, minted market cap, and fully diluted valuation; its FDV definition uses maximum supply multiplied by price ([CoinMarketCap methodology](https://support.coinmarketcap.com/hc/en-us/articles/360043836811-Market-Capitalization-Cryptoasset-Aggregate)).
Market cap asks: **What value does today’s reference price imply for the supply considered circulating?**
FDV asks: **What value would the same price imply if the selected full-supply measure were outstanding?**
FDV does **not** say the future network will be worth that amount. If supply changes, price, demand, protocol usage, and market conditions can change too.
**Scope and assumptions:** This guide addresses fungible cryptoassets. Stablecoins, rebasing tokens, wrapper tokens, algorithmic supply systems, uncapped assets, and assets with disputed circulation require additional treatment. All examples are hypothetical.
*This material is research, not financial advice. BlockMind will not tell you what to buy or sell, and its agent cannot execute transactions or move funds.*
## Methodology [#methodology]
**Observation date: July 10, 2026.** This comparison uses one timestamp and one reference-price methodology for each asset, keeps circulating, unlocked, total, and maximum supply separate, verifies who can change supply, and then adds a time-bucketed unlock schedule and standardized liquidity checks. It does not rank assets by a single ratio. This is the valuation layer inside the broader [pre-buy due diligence process](/blog/what-to-check-before-buying-crypto), not a standalone decision rule.
## Four supply numbers that should not be collapsed [#four-supply-numbers-that-should-not-be-collapsed]
### Circulating supply [#circulating-supply]
The amount a data provider considers available to the public market. This is a classification, not merely an on-chain total. Provider rules may exclude locked team tokens, treasury balances, or other restricted allocations.
### Unlocked supply [#unlocked-supply]
Tokens that are transferable under the relevant vesting or contractual restrictions. “Unlocked” does not necessarily mean circulating: a beneficiary may hold unlocked tokens without transferring them, and provider methodologies may classify treasury or insider balances differently.
### Total or minted supply [#total-or-minted-supply]
Tokens created, often net of verifiable burns. It may include locked and non-circulating balances.
### Maximum supply [#maximum-supply]
The stated upper bound, if one exists. Some systems have a fixed cap; others use governance-adjustable issuance or no fixed maximum. Avalanche’s builder documentation, for example, distinguishes fixed supply from elastic designs in which emissions, vesting, minting, or burning adjust supply ([Avalanche Builder Hub](https://docs.avax.network/academy/avalanche-l1/l1-native-tokenomics/01-tokens-fundamentals/02-token-design)).
Before comparing valuations, write down which supply belongs in each calculation and who can change it.
## The core formulas and what they mean [#the-core-formulas-and-what-they-mean]
Let:
* `P` = current reference price
* `C` = circulating supply
* `U` = unlocked supply
* `T` = total supply
* `M` = maximum supply
Then:
```text
Circulating market cap = P × C
Unlocked valuation = P × U
Minted valuation = P × T
FDV = P × M
```
### Circulation ratio [#circulation-ratio]
```text
Circulation ratio = C / M
```
If maximum supply is valid, the ratio shows how much of that full-supply denominator is currently classified as circulating.
### FDV-to-market-cap ratio [#fdv-to-market-cap-ratio]
```text
FDV / market cap = M / C
```
This identity holds when both calculations use the same price and the specified supply measures. It is a compact way to express the dilution gap, not its timing.
### Non-circulating gap [#non-circulating-gap]
```text
Non-circulating gap = M − C
Gap as % of current circulation = (M − C) / C × 100%
```
A large gap can be released over months, decades, conditionally, or never. That is why the calendar matters more than the ratio alone.
## Worked hypothetical: Token A and Token B [#worked-hypothetical-token-a-and-token-b]
**The assets and values below are fictional. They are not current market data, forecasts, or recommendations.**
Both tokens trade at a reference price of $2.
| Metric | Token A | Token B |
| ---------------------- | -----------: | -----------: |
| Circulating supply | 100 million | 100 million |
| Maximum supply | 1 billion | 200 million |
| Circulating market cap | $200 million | $200 million |
| FDV | $2 billion | $400 million |
| Circulation ratio | 10% | 50% |
They have the same circulating market cap, but Token A’s FDV-to-market-cap ratio is 10× and Token B’s is 2×.
That still does not establish which is “cheaper.” Now add the schedule:
* Token A: 30 million tokens unlock next year; the rest emits over 20 years for network security.
* Token B: 80 million investor and team tokens unlock next year; the remaining 20 million is a community reserve.
Token B has the smaller headline dilution gap but the larger near-term insider unlock. FDV alone hid the more decision-relevant timing.
Now add utility and demand:
* Token A is required for fees and security, but current usage is flat.
* Token B is a governance token with no required product payment, while protocol use is growing.
The conclusion remains conditional. Token A has structural utility but weak current demand; Token B has improving product use but unclear token value capture. Market cap and FDV organize the questions. They do not answer them.
## A better comparison: the dilution-gap worksheet [#a-better-comparison-the-dilution-gap-worksheet]
Use this six-part worksheet for each asset.
### 1. Verify the denominator [#1-verify-the-denominator]
| Question | Evidence |
| ----------------------------------------------------- | ----------------------------------------- |
| What does the provider count as circulating? | Methodology + project disclosure |
| Is maximum supply hard-coded, governed, or estimated? | Verified code + protocol docs |
| Are burns included and provable? | On-chain burn mechanism + balances |
| Are wrappers or bridged copies double-counted? | Token architecture and provider treatment |
If supply confidence is low, valuation precision is low.
### 2. Build a time-bucketed supply bridge [#2-build-a-time-bucketed-supply-bridge]
```text
Current circulating supply
+ expected net issuance
+ scheduled unlocks likely to become circulating
− scheduled burns
= estimated future circulating supply
```
Build it for 3, 6, 12, and 24 months. Keep “unlocked” separate from “assumed circulating.” Explain the assumption for each allocation.
The detailed process is in [Token Unlocks and Vesting](/blog/token-unlocks-vesting-guide).
### 3. Identify recipients and cost basis [#3-identify-recipients-and-cost-basis]
Supply entering validator rewards, a community program, a treasury, a market maker, an employee wallet, and an early investor wallet can have different incentives. Do not assume that every recipient sells, but do not ignore the recipient either.
Useful fields:
* Recipient class
* Tokens and percentage of current circulation
* Unlock mechanism
* Verifiable wallet or contract
* Disclosed acquisition terms, if public
* Historical transfer behavior, without inferring intent
### 4. Measure absorptive capacity [#4-measure-absorptive-capacity]
Compare the potential new circulation with actual liquidity, not just market cap:
* Median spot volume under a disclosed methodology
* Order-book depth and DEX price impact
* Number of credible venues
* Existing treasury or insider concentration
* Organic fee or usage demand
A $50 million unlock against a $1 billion market cap is not directly comparable with a $50 million market sell order. Market cap is price times supply, not available bids. Use [Crypto Liquidity Analysis](/blog/crypto-liquidity-analysis) to test realistic size.
### 5. Explain token value capture [#5-explain-token-value-capture]
Ask whether product success creates token demand or reduces supply:
* Is the token required for gas, collateral, staking, access, or governance?
* Are fees paid to holders, burned, retained by a company, or unrelated to the token?
* Can users access the product without holding the token?
* Can governance change issuance or utility?
“The protocol is growing” and “the token captures that growth” are separate claims.
### 6. Compare like with like [#6-compare-like-with-like]
Peer comparison requires aligned definitions:
* Same timestamp or sufficiently close market snapshot
* Same reference-price methodology
* Comparable supply category
* Same protocol category and token role
* Similar unlock horizon
* Similar revenue and fee definitions
Do not compare a base-layer gas asset, a stablecoin, and a governance token solely because all have a market cap.
## Scenario math without pretending to forecast [#scenario-math-without-pretending-to-forecast]
Scenario analysis can show arithmetic sensitivity if assumptions are explicit.
### Constant-market-cap dilution scenario [#constant-market-cap-dilution-scenario]
If circulating supply rises from `C0` to `C1` while circulating market cap remains exactly constant, then:
```text
P1 = P0 × C0 / C1
```
Suppose a fictional token has:
* `P0 = $4`
* `C0 = 100 million`
* `C1 = 125 million`
Under the **constant market cap assumption only**:
```text
P1 = $4 × 100 / 125 = $3.20
```
This is not a price target. It shows how the price would have to adjust if one variable—market cap—did not change. In reality, demand and valuation can rise or fall.
### Demand required to preserve price [#demand-required-to-preserve-price]
If price stays at `P0` as circulation increases by `ΔC`, the implied increase in circulating market cap is:
```text
Required valuation increase = P0 × ΔC
```
It is not necessarily the cash inflow required; market prices are set at the margin. Call it an **implied valuation increase**, not “new money needed.”
### Annualized net supply growth [#annualized-net-supply-growth]
```text
Net supply growth = (new circulation + emissions − burns) / starting circulation
```
Use projected and realized versions. A schedule can change, burns can depend on activity, and unlocked tokens may not enter measured circulation immediately.
## When is a high FDV-to-market-cap ratio concerning? [#when-is-a-high-fdv-to-market-cap-ratio-concerning]
It deserves more scrutiny when several conditions combine:
* A small public float sets the reference price
* Large near-term insider or investor unlocks
* Thin or concentrated liquidity
* Weak disclosed token demand
* Governance can expand supply
* Supply data is inconsistent or opaque
* FDV exceeds relevant peers without a clear mechanism-level reason
It can be less informative when:
* Most non-circulating supply releases slowly over decades
* Emissions pay for measurable network security
* Maximum supply is theoretical or governance-dependent
* Burns and issuance are activity-dependent
* The asset has no meaningful fixed maximum
The Optimism documentation illustrates why current sources matter: it describes the initial OP supply and points readers to an estimated circulating-supply schedule while governance can vote on token allocations and inflation ([Optimism docs](https://docs.optimism.io/governance/capital-allocation)). A static launch chart cannot substitute for current governance and supply data.
## Common market cap and FDV mistakes [#common-market-cap-and-fdv-mistakes]
### “A $1 token is cheaper than a $1,000 token” [#a-1-token-is-cheaper-than-a-1000-token]
Unit price ignores supply. Compare valuation, token role, liquidity, and economics.
### “FDV is the future market cap” [#fdv-is-the-future-market-cap]
FDV applies today’s price to a chosen full-supply denominator. Both can change.
### “Market cap is how much money entered the token” [#market-cap-is-how-much-money-entered-the-token]
It is a derived value, not accumulated deposits. Marginal trades can move the reference price applied to every circulating token.
### “All unlocks are immediate selling pressure” [#all-unlocks-are-immediate-selling-pressure]
Unlocking creates transferability or claimability, depending on design. It does not reveal holder intent. Verify transfers and market behavior after the event.
### “Max supply is always fixed” [#max-supply-is-always-fixed]
It can be absent, estimated, or governance-adjustable. Check code and governance.
### “Two provider market caps should match exactly” [#two-provider-market-caps-should-match-exactly]
They can use different prices, supply classifications, or update times. Investigate the definition rather than choosing the larger number.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Circulating supply is partly a methodology judgment.
* Reference prices aggregate venues with different quality and liquidity.
* Maximum supply may be undefined or changeable.
* FDV ignores time value and release schedule.
* Market cap and FDV do not measure protocol cash flow, security, governance quality, or token value capture.
* A high dilution gap does not prove future price decline; growing demand can offset supply.
* A low dilution gap does not prove safety; current holders can still be concentrated and liquid.
Use these metrics as the supply layer inside the broader [Crypto Market Analysis Guide](/blog/crypto-market-analysis-guide), not as standalone verdicts.
## A prompt for auditable analysis [#a-prompt-for-auditable-analysis]
Ask your BlockMind agent:
> “For \[asset], show circulating, unlocked, total, and maximum supply with definitions and primary sources. Calculate market cap and FDV from the same timestamp. Build a 3/6/12/24-month supply bridge, separate unlocks from assumed new circulation, identify recipients, and list every unresolved discrepancy.”
Your agent can combine current public research with [on-chain intelligence](/docs/capabilities/on-chain), but supply labels and future schedules still require human verification. Save the worksheet to your [Notebook](/docs/memory/notebook) and review it at each material governance or unlock event.
## The Bottom Line [#the-bottom-line]
Market cap describes today’s circulating-supply valuation under a provider’s definition. FDV applies today’s price to a fuller supply measure. The difference is a question generator, not a verdict.
Verify the denominator, map the release schedule, identify recipients, test liquidity, explain token value capture, and run transparent scenarios. If you cannot state the supply definitions and dates, the valuation comparison is not ready.
## Sources [#sources]
* [CoinMarketCap — Market Capitalization Methodology](https://support.coinmarketcap.com/hc/en-us/articles/360043836811-Market-Capitalization-Cryptoasset-Aggregate)
* [Avalanche Builder Hub — Token Design and Supply Models](https://docs.avax.network/academy/avalanche-l1/l1-native-tokenomics/01-tokens-fundamentals/02-token-design)
* [Avalanche Builder Hub — Vesting Schedules](https://docs.avax.network/academy/avalanche-l1/l1-native-tokenomics/07-token-distribution/02-vesting-schedules)
* [Optimism Documentation — Capital Allocation and OP Supply](https://docs.optimism.io/governance/capital-allocation)
* [CFTC — Use Caution When Buying Digital Coins or Tokens](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)
---
# Memecoin Due-Diligence Checklist: 20 Checks Before You Risk Capital
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Scams and Token Safety
Search intent: problem-solving
Primary keyword: memecoin due diligence checklist
Keywords: memecoin due diligence checklist, how to research memecoins, memecoin safety checklist, memecoin DYOR, check memecoin before buying
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Memecoin due diligence is less about forecasting utility and more about verifying control, exit conditions, incentives, and promotion. Start with the exact contract address, then check sellability, mutable permissions, liquidity ownership, holder concentration, insider overhang, and whether attention is authentic. A clean checklist reduces unknowns; it never proves safety or future value.
***
Memecoins are often driven primarily by culture, attention, and speculation rather than cash flow or functional utility. The SEC Division of Corporation Finance's [2025 staff statement](https://www.sec.gov/newsroom/speeches-statements/staff-statement-meme-coins) describes typical meme coins as inspired by memes or trends, often with limited or no functionality, significant volatility, and value driven mainly by market demand and speculation.
Use this checklist as the memecoin-specific companion to the cluster pillar, [Crypto Rug Pull Checker Guide](/blog/crypto-rug-pull-checker-guide). The pillar explains what automated tools inspect and what still requires manual evidence.
That statement is not a blanket safety or legal classification. It says the analysis depends on the facts, covers only coins consistent with its description, and is a staff view without the force of a Commission rule. It also notes that fraudulent conduct may be pursued under other laws. Commissioner Caroline Crenshaw published a [dissenting response](https://www.sec.gov/newsroom/speeches-statements/crenshaw-response-staff-statement-meme-coins-022725), arguing that the category is vague and that individual offers require facts-and-circumstances analysis. Both are relevant counterevidence to simplistic headlines.
## The hard-stop conditions [#the-hard-stop-conditions]
Pause before completing the rest of the checklist if any of these apply:
* The exact contract or mint address cannot be established.
* Independent simulations show ordinary holders cannot sell, or the result is unresolved.
* The token requests a seed phrase, private key, or unexplained signature merely to research it.
* An admin can mint, blacklist, pause, upgrade, or change taxes without a credible explanation and control structure.
* The claimed audit does not cover the deployed contract.
* A partnership or listing used as the main catalyst is denied or absent from the other party's official channel.
* The promoter guarantees returns or pressures immediate action.
Hard stop does not mean “fraud legally proven.” It means the current evidence is not strong enough to justify further exposure.
## 1. Verify the exact token identity [#1-verify-the-exact-token-identity]
* [ ] Full contract or mint address copied from an official source
* [ ] Chain and network confirmed
* [ ] Address cross-checked against an independent authoritative listing
* [ ] Main pool address and DEX version recorded
Names and symbols can be copied. Keep the full address in your research notes. If the project migrates contracts, treat the new token as a fresh investigation.
## 2. Check whether ordinary holders can sell [#2-check-whether-ordinary-holders-can-sell]
* [ ] Two independent simulation tools agree on sellability
* [ ] Buy, sell, and transfer taxes recorded
* [ ] Maximum sell and wallet limits understood
* [ ] Active pair used in the simulation matches the live market
* [ ] Result timestamp saved
[Honeypot.is](https://docs.honeypot.is/ishoneypot) documents sell simulation, taxes, gas, and maximum amounts, but also documents failed and indeterminate states. Use the four-state interpretation in [What Is a Honeypot Crypto Token?](/blog/honeypot-crypto-token). Do not spend real money merely to test suspicious code.
## 3. Map contract powers [#3-map-contract-powers]
* [ ] Source code or program is verified where the chain supports it
* [ ] Proxy and current implementation identified
* [ ] Owner and every privileged role identified
* [ ] Mint, pause, blacklist, whitelist, tax, limit, and upgrade powers recorded
* [ ] Timelock or multisig control verified on-chain
* [ ] Recent administrative events reviewed
“Ownership renounced” is only one field. OpenZeppelin documents role-based access control and contract ownership, so separate minter, pauser, upgrader, and administrator roles may survive.
## 4. Verify liquidity and who can remove it [#4-verify-liquidity-and-who-can-remove-it]
* [ ] Every material pool listed
* [ ] Executable depth tested at realistic sizes
* [ ] LP-token holders or position NFTs identified
* [ ] Locker contract verified through the locker's official site
* [ ] Locked percentage and exact unlock time calculated
* [ ] Active concentrated-liquidity range checked
* [ ] Unlocked positions and alternative pools reviewed
Use the full [liquidity locks guide](/blog/liquidity-locks-crypto-explained). A lock is meaningful evidence about a specified position. It does not prove the token is safe.
## 5. Analyze holder concentration [#5-analyze-holder-concentration]
* [ ] Pools, bridges, treasuries, burns, lockers, vesting contracts, and exchanges labeled
* [ ] Top non-system wallets measured as a share of circulating supply
* [ ] Deployer funding paths traced
* [ ] Related transfers and coordinated timing reviewed
* [ ] Dust/airdrop holders excluded from “community” claims
* [ ] Team and promoter wallets reconciled to disclosures
One person can control many addresses. Look for shared funding sources, synchronized transactions, and direct deployer links without claiming identity where evidence stops.
## 6. Calculate liquidity-adjusted insider overhang [#6-calculate-liquidity-adjusted-insider-overhang]
Market-cap percentage alone understates exit risk. Compare potentially liquid insider holdings with actual exit depth:
`insider-overhang ratio = value of unlocked, linked holdings ÷ executable pool depth within chosen price impact`.
**Hypothetical example:** Deployer-linked wallets hold tokens marked at $600,000. Aggregate depth available within a 10% price-impact band is $75,000. The ratio is `600,000 ÷ 75,000 = 8`.
That does not predict an eightfold loss or prove the wallets will sell. It shows that marked insider holdings are eight times the near-price depth used in the calculation. Distributing even part of them could overwhelm the current market.
Record the data source, block/time, chosen price band, and wallet-label rationale so someone else can reproduce the estimate.
## 7. Inspect token supply and unlocks [#7-inspect-token-supply-and-unlocks]
* [ ] Circulating, total, and maximum supply definitions reconciled
* [ ] Mint authority and emissions understood
* [ ] Team, promoter, treasury, market-maker, and airdrop allocations mapped
* [ ] Vesting contracts and cliffs verified on-chain
* [ ] Burns verified by transaction and address, not announcement
* [ ] Upcoming unlocks compared with current depth and volume
Our [tokenomics research guide](/blog/how-to-research-crypto-tokenomics) explains the supply terms. For a memecoin with no functional demand, dilution and insider distribution can dominate the economics.
## 8. Verify the team and promoter claims [#8-verify-the-team-and-promoter-claims]
* [ ] Team identity claims cross-checked where identities are public
* [ ] Prior projects and exits researched
* [ ] Promoter holdings and compensation disclosed
* [ ] Partnerships verified on the partner's domain
* [ ] Exchange listings verified on the exchange's official channel
* [ ] Audit opened on the auditor's domain
* [ ] Audit address, chain, commit, scope, and unresolved findings matched
Pseudonymous teams are not automatically fraudulent. They reduce an accountability layer, so technical control, disclosures, and operational evidence need to be stronger.
## 9. Evaluate attention quality [#9-evaluate-attention-quality]
* [ ] Earliest source of the narrative identified
* [ ] Engagement checked for repetitive, generic, or bot-like replies
* [ ] Follower growth compared with product or market events
* [ ] Giveaways and paid promotion separated from organic discussion
* [ ] Critical questions and moderation behavior reviewed
* [ ] Countdown groups, guaranteed returns, and “do not sell” instructions flagged
The CFTC warns against buying on a single social-media tip, false urgency, or sudden spike. Social attention is a market input, not verification of underlying claims.
## 10. Compare market activity with economic reality [#10-compare-market-activity-with-economic-reality]
* [ ] Volume compared across independent venues and data providers
* [ ] Repetitive trade sizes and rapid round trips investigated
* [ ] Active traders traced for common funding where feasible
* [ ] Price move matched to a primary-source catalyst
* [ ] Large holder and deployer flows reviewed during the move
* [ ] Liquidity additions and removals overlaid on price and promotion
Chainalysis' [2025 methodology](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/) found patterns consistent with wash trading and pump-and-dumps, but it explicitly cautions that on-chain patterns alone do not prove intent. Use the [pump-and-dump signs guide](/blog/crypto-pump-and-dump-signs) to maintain legitimate alternative explanations.
## 11. Test the cultural thesis honestly [#11-test-the-cultural-thesis-honestly]
Because many memecoins have little functional utility, the thesis may be “this community and symbol will retain attention.” Make that claim falsifiable:
* What audience exists beyond price speculation?
* Is the meme original, durable, and legally usable—or copied from a temporary event?
* Who controls the official accounts, domain, treasury, and brand assets?
* Can the promoter change the narrative or abandon distribution channels?
* What would show attention is deteriorating?
* Is the market already pricing an implausibly large share of future attention?
A cultural thesis can be real without being measurable like protocol revenue. That makes position sizing and exit liquidity more—not less—important, but those choices are personal financial decisions outside this research guide.
## 12. Keep an auditable evidence ledger [#12-keep-an-auditable-evidence-ledger]
Use one row per claim:
| Claim | Primary evidence | Timestamp/block | Status | Counterevidence | Recheck trigger |
| ------------------------- | ----------------------------- | --------------- | ------------------------------ | -------------------------- | -------------------- |
| Ordinary holders can sell | Two simulations on named pool | Record it | Pass/fail/unknown | Mutable tax role | Role or proxy change |
| Main liquidity is locked | Locker contract + position | Record it | Full/partial/unknown | Other unlocked pool | Unlock or migration |
| Supply is fixed | Implementation + all roles | Record it | Verified/false/unknown | Upgrade authority | Role grant/upgrade |
| Promoter is not selling | Disclosed wallets + flows | Record it | Supported/contradicted/unknown | Undisclosed linked wallets | Large transfer |
“Unknown” is a valid and important status. Do not replace it with a guess.
## 13. Set monitoring triggers before the hype [#13-set-monitoring-triggers-before-the-hype]
Research can become stale within a block. Useful recheck triggers include:
* proxy upgrade or privileged-role change;
* tax, blacklist, pause, or trading-setting change;
* liquidity unlock, migration, or large removal;
* deployer-linked transfer;
* holder-concentration change;
* new audit or exploit disclosure;
* exchange-listing claim;
* abnormal volume or social spike;
* contract migration.
For a wider alert design, see [Crypto Alerts Beyond Price](/blog/crypto-alerts-beyond-price). Alerts are signals to investigate, not instructions to trade.
## A compact 20-check printout [#a-compact-20-check-printout]
1. Exact address verified
2. Chain verified
3. Active pool verified
4. Two sell simulations run
5. Taxes and limits recorded
6. Source/implementation verified
7. All privileged roles mapped
8. Mint and upgrade powers understood
9. Main liquidity owners mapped
10. Lock amount and date verified
11. Exit depth tested
12. System holder addresses labeled
13. Deployer-linked wallets traced
14. Insider overhang calculated
15. Supply and unlocks reconciled
16. Audit scope matched
17. Partnerships/listings independently confirmed
18. Promoter compensation and holdings checked
19. Volume and social activity investigated
20. Evidence ledger and recheck triggers saved
Completing all 20 does not make the token safe. It makes your remaining uncertainty visible.
## Limitations [#limitations]
Public data cannot reveal every beneficial owner, private agreement, compromised key, or planned action. Simulation tools support limited chains and routes. Liquidity and price-impact data change. An audit can miss vulnerabilities. A real community can still disappear, and a transparent token can still lose most of its value.
Legal treatment varies by token, transaction, facts, and jurisdiction. The SEC staff statement discussed above has explicit limits and dissenting views; do not treat it as personalized legal advice.
## How BlockMind fits [#how-blockmind-fits]
A BlockMind agent can help execute the research workflow: open primary sources, organize on-chain observations, maintain the evidence ledger in your [Notebook](/docs/memory/notebook), and periodically monitor a tracked idea. A free [DeepDive report](https://blockmind.app/deepdive) can provide a structured first pass.
Neither proves safety. The agent can be wrong or stale, cannot identify every hidden owner, never trades, and never tells you what to buy or sell.
## The Bottom Line [#the-bottom-line]
Memecoin due diligence cannot manufacture fundamentals that do not exist. It can reveal who controls the token, whether ordinary holders can exit, how much liquidity supports the marked value, and whether promotion matches observable behavior.
Use the checklist to make unknowns explicit. If a hard-stop condition remains unresolved, a green scanner badge or viral meme is not contrary evidence.
This article is for research and education, not financial, legal, or tax advice.
## Sources [#sources]
1. [SEC Division of Corporation Finance: Staff statement on meme coins](https://www.sec.gov/newsroom/speeches-statements/staff-statement-meme-coins)
2. [SEC Commissioner Crenshaw: Response and counterview on the staff statement](https://www.sec.gov/newsroom/speeches-statements/crenshaw-response-staff-statement-meme-coins-022725)
3. [CFTC: Use caution when buying digital coins or tokens](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/caution_of_digital_currencies.html)
4. [CFTC: Beware virtual-currency pump-and-dump schemes](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/beware_virtual_currency_pump_dump.html)
5. [Honeypot.is: Sell-simulation documentation](https://docs.honeypot.is/ishoneypot)
6. [OpenZeppelin: Smart-contract access control](https://docs.openzeppelin.com/contracts/5.x/access-control)
7. [Uniswap: Liquidity-position ownership](https://support.uniswap.org/hc/en-us/articles/20980786685069-Why-is-liquidity-position-ownership-represented-by-tokens-or-NFTs)
8. [Chainalysis: Market-manipulation research and methodological limits](https://www.chainalysis.com/blog/crypto-market-manipulation-wash-trading-pump-and-dump-2025/)
---
# Read-Only Crypto API Keys: What They Protect and How to Verify Permissions
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Portfolio Monitoring and Risk
Search intent: problem-solving
Primary keyword: read-only crypto API keys
Keywords: read-only crypto API keys, crypto API key security, portfolio tracker API key safety, exchange read-only API, API key permissions crypto
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A provider-scoped read-only exchange API key can expose balances and account history but cannot place trades or withdraw funds. That sharply reduces financial authority, not all risk. Verify the exchange's own permission screen, disable every write scope, protect the credential like a password, and revoke it when no longer needed or after suspected exposure.
***
Portfolio trackers need data, not control. The safest exchange connection is therefore the narrowest credential that can read the required balances and transactions without trading, transferring, withdrawing, or changing account settings.
This is part of the wider [AI portfolio monitoring](/blog/ai-portfolio-monitoring-crypto) trust boundary. Monitoring becomes more useful with current holdings, but portfolio context should not require execution authority.
## What is a read-only crypto API key? [#what-is-a-read-only-crypto-api-key]
An API key is a credential an application uses to authenticate requests to an exchange. “Read-only” describes the permissions attached to that credential, not the secrecy of the key itself.
A properly scoped read-only key may be able to retrieve:
* balances;
* positions and margin status;
* deposits and withdrawals history;
* orders and trades history;
* ledgers and transaction history;
* account or portfolio identifiers.
It should not be able to:
* place, edit, or cancel orders;
* transfer assets between accounts where that is a write operation;
* add withdrawal addresses;
* withdraw assets;
* change security or user settings.
The precise boundary is provider-specific. Trust the exchange's current permission documentation and the key-creation screen, not a generic tutorial.
## Provider permission matrix [#provider-permission-matrix]
Checked against official documentation on July 10, 2026:
| Provider | Read scope | Trade/write scope to leave off | Withdrawal/transfer scope to leave off | Useful security control |
| --------------------- | -------------------------------------------- | ------------------------------------------------------ | --------------------------------------------- | ----------------------------------------------- |
| **Coinbase Exchange** | `View` | `Trade`, `Manage` | `Transfer` | Required IP allowlist for Exchange API keys |
| **Kraken** | Query funds, orders/trades, ledger as needed | Modify/close trades, earn or other unnecessary actions | Withdraw funds; add/update withdrawal address | Permission-specific keys; API-key 2FA available |
| **OKX** | `Read` | `Trade` | `Withdraw` | IP address binding |
| **Binance** | Account/user-data access needed by tracker | `TRADE` and other write permissions | Withdrawal-related authority | Endpoint-type restrictions and IP allowlist |
Coinbase's [permission documentation](https://help.coinbase.com/en/exchange/managing-my-account/how-to-create-an-api-key) is particularly explicit: `View` reads endpoints, `Trade` posts orders, and `Transfer` can transfer value, including withdrawals. Kraken exposes separate query, trade-modification, and withdrawal permissions. OKX separates Read, Trade, and Withdraw. Binance says keys can be restricted by secure endpoint type and trading is not enabled by default.
Interfaces and names can change. Recheck the official page while creating the key.
## The READ verification protocol [#the-read-verification-protocol]
### R — Requirements first [#r--requirements-first]
Write down what the application needs before opening the exchange settings:
* balances only;
* balances plus transaction history;
* open positions;
* tax lots or ledger history;
* specific portfolio or subaccount;
* sync frequency;
* source IPs, if allowlisting is supported.
If the use case is portfolio analysis, trade and withdrawal authority are not requirements.
### E — Exchange-native scope [#e--exchange-native-scope]
Create the key inside the exchange's official authenticated interface. Do not follow links from unsolicited messages or search ads. Verify domain, account, and subaccount.
Select only read/query permissions. Some exchanges separate several query scopes; enable only those needed for the documented import. Take a screenshot or write down the final scope without capturing the secret.
### A — Access controls and application trust [#a--access-controls-and-application-trust]
* Enable strong MFA on the exchange and connected application.
* Use IP allowlisting when the application's documented infrastructure supports stable addresses.
* Create a separate key per application instead of reusing one.
* Give the key a descriptive name and creation date.
* Check how the application stores and processes credentials.
* Confirm how to delete imported data and revoke access.
* Never paste API secrets into chat, email, tickets, source code, or shared documents.
Kraken's [API security guidance](https://support.kraken.com/hc/articles/api-key-security) recommends encrypted storage or a reputable password manager, minimal permissions, API-key 2FA where available, and deleting keys that are no longer relevant. NIST's least-privilege principle likewise says processes should receive only the access necessary for assigned tasks.
### D — Detect and decommission [#d--detect-and-decommission]
After connection:
* confirm the app shows the intended account and no others;
* review key usage where the exchange exposes it;
* inspect exchange activity and security notifications;
* rotate or recreate the key according to your risk policy and provider guidance;
* revoke it before deleting the application account, when no longer needed, or immediately after suspected exposure;
* verify that revocation breaks the sync.
Deleting a portfolio inside an app may not revoke the exchange key. Delete it at the exchange.
## The read-only proof test [#the-read-only-proof-test]
Do not accept a “read-only” label until all four statements are true:
1. The exchange permission screen shows no order-placement or trade-modification scope.
2. It shows no transfer, withdrawal, or withdrawal-address-management scope.
3. It shows no account-security or broad management scope.
4. The app's documented use case requires only the selected query permissions.
If a provider offers one combined permission that includes actions beyond the use case, it is not provider-scoped read-only even if the application promises to behave read-only.
## BlockMind's connection boundary—and the Firi exception [#blockminds-connection-boundaryand-the-firi-exception]
For Binance, Coinbase, Kraken, OKX, KuCoin, MEXC, Crypto.com, and Bitstamp, BlockMind instructs users to create provider-scoped read-only keys with trading and withdrawals disabled. BlockMind uses these connections only to read balances and positions and cannot initiate trades, withdrawals, or transfers.
**Firi is the explicit exception:** Firi does not offer a provider-scoped read-only API key. BlockMind stores the provided credentials encrypted and enforces balance-only behavior in the application, but the exchange itself does not restrict that key to reading. That means the credential deserves greater caution and faster revocation after suspected compromise.
The canonical provider list and current setup details live in [Connect your portfolio](/docs/getting-started/connect-your-portfolio). The broader data-handling boundary is in [Trust and security](/docs/security).
## Read-only does not mean risk-free [#read-only-does-not-mean-risk-free]
### Privacy risk [#privacy-risk]
Balances, positions, trade history, deposits, withdrawals, and account relationships are sensitive. They can reveal net worth, strategies, counterparties, and behavior.
### Credential risk [#credential-risk]
An API secret is still a secret. A service compromise, phishing attack, browser extension, malware, log, screenshot, or exposed support ticket can leak it. Misconfigured permissions turn the same leaked credential into a financial threat.
### Data-integrity risk [#data-integrity-risk]
Read-only access prevents the app from writing to the exchange. It does not guarantee that imported balances, transaction history, cost basis, or asset mappings are complete. Exchange APIs can omit product types or historical activity.
### Account-linking risk [#account-linking-risk]
Combining wallets and exchanges in one tracker creates a high-value map of holdings. Review the provider's privacy, retention, deletion, and breach practices.
### Social-engineering risk [#social-engineering-risk]
Scammers may impersonate support and ask for a key, QR code, remote screen access, seed phrase, or “verification” payment. A legitimate portfolio connection never requires your wallet seed phrase or private key.
## A blast-radius worksheet [#a-blast-radius-worksheet]
Inventory one row per credential:
| Key | Account scope | Read data visible | Write scopes | IP restriction | Created | Last used | Revoke trigger |
| ---------- | ------------------ | ----------------- | ------------ | -------------- | ------- | --------- | --------------------------- |
| Tracker A | Spot portfolio | Balances + ledger | None | Yes/no | Date | Date | App removal or exposure |
| Tax tool B | Main account | Full history | None | Yes/no | Date | Date | Filing complete/review date |
| Bot C | Trading subaccount | Balances + orders | Trade | Yes/no | Date | Date | Strategy stop or anomaly |
### Worked hypothetical [#worked-hypothetical]
A user has three services connected to the same exchange:
* Portfolio tracker: read balances and ledger.
* Tax tool: read full transaction history.
* Old bot: read and trade.
The user stopped using the bot six months ago but left its key enabled. The portfolio tracker and tax tool cannot trade; the dormant bot can. The largest preventable blast radius comes from the unused trade-enabled key, not the two active read-only keys. The correct security task is credential inventory and revocation—not disconnecting the useful read-only data first.
This worksheet is qualitative. It does not estimate breach probability.
## API key vs wallet address vs seed phrase [#api-key-vs-wallet-address-vs-seed-phrase]
| Credential/data | What it normally enables | Safe for a balance tracker? |
| --------------------------------- | ---------------------------------- | ------------------------------------- |
| Public wallet address | View public on-chain activity | Yes, with privacy tradeoffs |
| Provider-scoped read-only API key | View permitted exchange data | Appropriate when required and trusted |
| Trade-enabled API key | Place or modify orders | Not needed for portfolio tracking |
| Withdrawal-enabled API key | Move assets | Never needed for portfolio tracking |
| Seed phrase/private key | Control and sign for wallet assets | Never share with a tracker |
If an application asks for a seed phrase to “import balances,” stop.
## Connection and revocation checklist [#connection-and-revocation-checklist]
* [ ] Official exchange domain and account confirmed
* [ ] Use case and required query fields written down
* [ ] Separate key created for this application
* [ ] Trade disabled
* [ ] Transfers and withdrawals disabled
* [ ] Address management and broad management disabled
* [ ] Scope limited to correct portfolio/subaccount where supported
* [ ] IP allowlist configured if compatible
* [ ] MFA enabled on exchange and app
* [ ] Secret stored only through the intended secure entry path
* [ ] First sync reconciled against source account
* [ ] Key added to credential inventory
* [ ] Revocation tested or procedure documented
* [ ] Review/revocation trigger scheduled
For tracker selection, compare how providers connect and reconcile data in [Best Crypto Portfolio Trackers](/blog/best-crypto-portfolio-trackers). For a full data-quality review after connection, use the [crypto portfolio health check](/blog/crypto-portfolio-health-check).
## Limitations and counterevidence [#limitations-and-counterevidence]
Provider documentation and interfaces change. Some connections use OAuth rather than an API key, and OAuth scopes require the same least-privilege review. IP allowlisting may not work with a service that does not publish stable egress addresses. A provider's “read” permission may expose more account history than you expect.
Read-only reduces authority; it does not guarantee the third party's security, privacy, accuracy, or availability. Evaluate the application as well as the key.
## The Bottom Line [#the-bottom-line]
Read-only crypto API keys are the right default for portfolio tracking because the job is observation, not execution. Prove the scope in the exchange UI, protect the secret, inventory every connection, and remove dormant credentials.
The decisive question is not “Does the app call itself read-only?” It is “What actions will the exchange authorize for this exact key?”
This article is for research and education, not personalized security or financial advice.
## Sources [#sources]
1. [Coinbase Exchange: API key permissions](https://help.coinbase.com/en/exchange/managing-my-account/how-to-create-an-api-key)
2. [Kraken: API key permission reference](https://support.kraken.com/articles/360000919966-how-to-create-an-api-key)
3. [Kraken: API key security practices](https://support.kraken.com/hc/articles/api-key-security)
4. [OKX: API key permissions and IP binding](https://my.okx.com/docs-v5/)
5. [Binance: API request security and endpoint permissions](https://developers.binance.com/legacy-docs/binance-spot-api-docs/rest-api/request-security)
6. [NIST SP 800-171 Rev. 3: Least privilege](https://nvlpubs.nist.gov/nistpubs/SpecialPublications/800-171r3/NIST.SP.800-171r3.html)
7. [BlockMind: Connect your portfolio](/docs/getting-started/connect-your-portfolio)
8. [BlockMind: Trust and security](/docs/security)
---
# Support and Resistance in Crypto: How to Draw and Test Zones
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: support and resistance crypto explained
Keywords: support and resistance crypto explained, crypto support and resistance, how to draw support and resistance crypto, crypto price zones, support resistance levels
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Support and resistance are zones where past order flow or trader attention suggests price may pause, reverse, or accelerate after a break. They are not guaranteed floors or ceilings. Draw them from a fixed market, timeframe, and historical window; define zone width and invalidation before observing the next move; and evaluate them with volume, volatility, liquidity, and costs.
Support lies below the current price and represents an area where buying or reduced selling may slow a decline. Resistance lies above price and represents an area where selling or reduced buying may slow a rise. Once price crosses a zone, traders often watch it for a possible role reversal—but “often watched” is not the same as “must hold.”
Technical structure is one layer in the wider [crypto market analysis guide](/blog/crypto-market-analysis-guide); it should be checked against liquidity, on-chain, sentiment, and event evidence.
For two complementary regime inputs, read the [Crypto Fear & Greed Index guide](/blog/fear-greed-index-guide) and [Bitcoin dominance explained](/blog/bitcoin-dominance-explained). Neither is a trade trigger either.
CME’s educational material defines support and resistance as levels where price may slow or reverse and lists previous highs/lows, key prices, moving averages, and trend lines as common methods ([CME Group](https://www.cmegroup.com/education/courses/trading-and-analysis/support-and-resistance.hideSubnav.educationIframe.html?hideAddThisExt=y\&hideFooter=y\&hideHeader=y\&hideRightRail=y)).
## Why zones can matter [#why-zones-can-matter]
### Orders cluster [#orders-cluster]
Participants place entries, exits, take-profits, and stop orders around memorable prices and prior turning areas. Federal Reserve Bank of New York research using order data found take-profit and stop-loss orders clustered at round numbers in foreign exchange and offered a microstructure explanation for reversals near levels and momentum after breaks ([New York Fed](https://www.newyorkfed.org/research/staff_reports/sr125.html)). This evidence is not crypto-specific, but the mechanism is plausible in any order-driven market.
### Market participants remember reference prices [#market-participants-remember-reference-prices]
Previous highs, lows, entry areas, and liquidation points can affect behavior when revisited. The response changes as participants and positions change.
### Many people watch similar charts [#many-people-watch-similar-charts]
Widely observed areas may become self-reinforcing for a time. They can also become attractive locations for stop runs and false breakouts.
### Liquidity is uneven [#liquidity-is-uneven]
Visible and hidden orders are not distributed uniformly. Thin books can pass through a chart level with little resistance; deeper liquidity may absorb a move.
## Why levels fail [#why-levels-fail]
* New information changes fair-value expectations.
* A large market order consumes available liquidity.
* Leverage liquidations create forced flow.
* The chosen timeframe is irrelevant to the active participants.
* The level was drawn after the outcome and only appears predictive in hindsight.
* The asset or venue has too little reliable volume.
* The zone was defined as a precise line despite ordinary volatility.
A level failing is not an anomaly. It is one of the expected outcomes.
## Five common ways to identify zones [#five-common-ways-to-identify-zones]
### 1. Previous swing highs and lows [#1-previous-swing-highs-and-lows]
Mark areas where price clearly changed direction. Use a rule for what counts as a swing, such as a local high/low separated by a minimum number of bars or volatility-adjusted move. Otherwise every wiggle becomes a level.
### 2. Repeated reaction areas [#2-repeated-reaction-areas]
Cluster nearby turning prices rather than drawing one line through an arbitrary wick. More reactions can increase attention, but repeated tests may also consume resting liquidity.
### 3. Consolidation boundaries [#3-consolidation-boundaries]
The upper and lower edges of a range show where price repeatedly failed to continue. Record whether boundaries are based on closes, bodies, or wicks.
### 4. Volume-based areas [#4-volume-based-areas]
Volume profiles and anchored volume-weighted average price can identify where substantial trading occurred. Results depend on venue coverage and the anchor. Crypto volume is fragmented across venues, so one chart may not represent the whole market.
### 5. Dynamic references [#5-dynamic-references]
Moving averages and trend lines move with time. They may act as shared reference points, but using many lengths until one fits is data snooping.
Fibonacci retracements and round numbers are also common. Treat them as hypotheses to test, not natural laws. If several methods produce nearby values, define one confluence zone before the next observation rather than adding methods after price reacts.
## The predeclared-zone protocol [#the-predeclared-zone-protocol]
This method makes the analysis auditable.
### Step 1: Freeze the market definition [#step-1-freeze-the-market-definition]
Record:
* asset and contract if relevant;
* spot or derivative instrument;
* venue or index;
* quote currency;
* candle interval and timezone;
* historical data range;
* price type and adjustment method.
BTC/USD spot on one venue is not the same series as a perpetual index.
### Step 2: Choose one identification method [#step-2-choose-one-identification-method]
Write the rule before looking forward. Example:
> Use daily candles. Identify swing lows with two lower lows on each side, cluster prices within 1.0 average true range, and require at least two historical reactions separated by seven days.
The numbers are illustrative, not recommended defaults. The important feature is precommitment.
### Step 3: Define zone width [#step-3-define-zone-width]
Crypto volatility makes exact lines brittle. Set width by:
* percentage of price;
* fraction of average true range;
* historical distribution of reaction distances; or
* observed liquidity band.
Do not widen the zone after price barely misses it.
### Step 4: Define a valid reaction [#step-4-define-a-valid-reaction]
Examples:
* price enters the zone and closes away by a chosen amount;
* price closes through, then reclaims within a fixed number of bars;
* the move occurs with a stated volume or order-flow condition.
“It touched and later went up” is too vague.
### Step 5: Define break and invalidation [#step-5-define-break-and-invalidation]
Specify whether a wick, close, consecutive closes, or volatility-adjusted distance counts as a break. Also define when the zone expires because time, regime, or structure changed.
### Step 6: Evaluate prospectively [#step-6-evaluate-prospectively]
Track every zone produced by the rule, not only memorable successes. Compare with a simple baseline and include fees, spread, slippage, and latency if evaluating a strategy.
## How to read price at a zone [#how-to-read-price-at-a-zone]
### Approach speed [#approach-speed]
A fast move can reflect urgency and may either break the zone or exhaust into it. The direction is not predetermined.
### Volume and order flow [#volume-and-order-flow]
Rising volume can confirm participation but does not tell you whether absorption or continuation wins. Look at how price responds to the volume.
### Volatility [#volatility]
When volatility expands, an old narrow zone may no longer be meaningful. Normalize distances.
### Liquidity and venue agreement [#liquidity-and-venue-agreement]
Check whether the move appears across major venues and whether visible depth can support the proposed interpretation. A wick on one thin market may be venue-specific.
### Higher-timeframe context [#higher-timeframe-context]
A daily zone usually matters more to a multi-week thesis than a five-minute line. CME’s timeframe guidance notes that daily charts can identify longer-term support and resistance ([CME Group](https://www.cmegroup.com/education/courses/using-the-trading-simulator/trading-sim-in-30-seconds-chart-time-frames)). Do not mix timeframes without stating which decision each serves.
### Catalyst and broader market [#catalyst-and-broader-market]
An exploit, listing, macro event, or Bitcoin-wide move can overwhelm asset-specific chart structure. Technical context is one evidence layer.
## Support/resistance role reversal [#supportresistance-role-reversal]
After a convincing break above resistance, traders may watch the old zone as support; after a break below support, as resistance. This can happen because trapped participants exit near their original reference price and breakout participants enter on a retest.
Do not label role reversal until you define the break and retest. If price oscillates through the area repeatedly, the zone may be losing informational value.
## Evidence: useful, but not universal [#evidence-useful-but-not-universal]
A New York Fed study of levels published by six foreign-exchange firms found that they helped predict intraday trend interruptions in its 1996–1998 sample, with performance varying by currency and firm ([Federal Reserve Bank of New York](https://www.newyorkfed.org/medialibrary/media/research/epr/00v06n2/0007osle.pdf)). A 2019 study of high-frequency Bitcoin returns tested technical rules, including trading-range breakouts, and found stronger support for moving-average strategies in its sample ([Finance Research Letters](https://doi.org/10.1016/j.frl.2019.04.027)).
These studies do not prove that a line drawn today will work, nor that results survive new data and trading costs. They support treating technical levels as testable hypotheses rather than either magic or nonsense.
## How AI can help without inventing certainty [#how-ai-can-help-without-inventing-certainty]
AI can:
* calculate zones from a declared rule;
* preserve chart, venue, timeframe, and data range;
* compare multiple methods without hiding disagreement;
* measure historical reaction and break rates;
* attach relevant volume, volatility, and news context;
* monitor a predeclared zone;
* explain what would invalidate the setup.
Use this prompt:
> On \[venue/pair/timeframe], identify support and resistance using \[method] and data ending at \[timestamp]. Define zone width before evaluating later candles. Show every qualifying zone, not selected examples. Report reactions, breaks, false breaks, and costs. Treat the output as context, not a trade signal.
BlockMind’s research journey can include technical levels in a wider analysis: [Explore → Analyze → Verdicts → Track](/docs/capabilities/research-journey). The agent still never tells you what to buy or sell.
Before relying on an AI-generated level, apply the [AI-analysis verification protocol](/blog/how-to-verify-ai-crypto-analysis) to its data window, venue, calculation, and citations.
## Common mistakes [#common-mistakes]
* drawing too many levels until price is always “near” one;
* using a single venue for an asset with fragmented liquidity;
* switching between wicks and closes after the result;
* ignoring timeframe and volatility;
* calling the first touch confirmation;
* moving the line to preserve the story;
* judging only successful examples;
* entering without considering spread and slippage;
* treating a level as a substitute for fundamental or on-chain research.
## Limitations and counterevidence [#limitations-and-counterevidence]
Technical zones can arise in random-looking series, and human pattern recognition is powerful enough to find them after the fact. Multiple analysts also choose different levels from the same chart. Predeclared rules and prospective tracking reduce—but do not eliminate—subjectivity and overfitting.
Visible support can attract stop placement just beyond the zone, making it a target. In illiquid tokens, manipulators can create or break apparent levels. No historical zone limits losses unless the investor independently manages exposure.
## The Bottom Line [#the-bottom-line]
Support and resistance are conditional map features, not instructions. Define the market, method, zone width, reaction, break, and expiry before evaluating the future. Use the zones with liquidity, volatility, volume, event, and portfolio context, and keep every failed level in the record.
This is research, not financial advice. BlockMind’s agent cannot trade or touch funds and never tells you what to buy or sell.
## Sources [#sources]
* [CME Group — Support and Resistance](https://www.cmegroup.com/education/courses/trading-and-analysis/support-and-resistance.hideSubnav.educationIframe.html?hideAddThisExt=y\&hideFooter=y\&hideHeader=y\&hideRightRail=y)
* [Federal Reserve Bank of New York — Support for Resistance](https://www.newyorkfed.org/medialibrary/media/research/epr/00v06n2/0007osle.pdf)
* [Federal Reserve Bank of New York — Currency Orders and Exchange-Rate Dynamics](https://www.newyorkfed.org/research/staff_reports/sr125.html)
* [Finance Research Letters — Effectiveness of Technical Trading Rules in Cryptocurrency Markets](https://doi.org/10.1016/j.frl.2019.04.027)
* [CME Group — Chart Time Frames](https://www.cmegroup.com/education/courses/using-the-trading-simulator/trading-sim-in-30-seconds-chart-time-frames)
---
# Crypto Token Unlocks and Vesting: How to Measure the Real Supply Event
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: informational
Primary keyword: token unlocks and vesting
Keywords: crypto token unlocks, token vesting schedule, token unlock calendar, cliff vesting crypto, crypto dilution analysis
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A token unlock is not the same as new circulating supply, and new circulating supply is not the same as selling. Analyze five separate stages: scheduled vesting, on-chain claimability, transferability, provider classification as circulating, and observed transfers to market venues. Measure the event against current circulation and executable liquidity, then document a range instead of assuming every unlocked token is sold.
***
## What is a crypto token unlock? [#what-is-a-crypto-token-unlock]
A crypto token unlock is the point at which tokens subject to a time, milestone, contract, or legal restriction become claimable or transferable under the applicable schedule. The exact event depends on the implementation.
Avalanche’s builder documentation describes the common components: total allocation, cliff, vesting period, and release interval, with linear, graded, and cliff-based schedules among the possible designs ([Avalanche Builder Hub](https://docs.avax.network/academy/avalanche-l1/l1-native-tokenomics/07-token-distribution/02-vesting-schedules)).
An unlock can mean:
* Tokens automatically transfer to a beneficiary
* A beneficiary becomes able to claim vested tokens
* A contractual restriction expires while tokens were always on-chain
* A timelock releases control of a treasury allocation
* A milestone makes a tranche eligible, subject to another action
Never rely on the word “unlock” without identifying which mechanism applies.
**Scope and assumptions:** This guide covers fungible-token vesting and scheduled distribution. It does not interpret employment agreements, SAFTs, securities law, tax obligations, or private side letters. Where the on-chain implementation and legal agreement differ, both may matter.
*BlockMind provides research, not financial advice. Its agent will not tell you what to buy or sell and cannot trade, withdraw, or move funds.*
Unlock analysis is one part of the complete [pre-buy crypto due diligence process](/blog/what-to-check-before-buying-crypto), not a standalone signal.
## The five stages most unlock calendars collapse [#the-five-stages-most-unlock-calendars-collapse]
### 1. Scheduled to vest [#1-scheduled-to-vest]
The published schedule says some amount should become vested on a date or over an interval.
### 2. Claimable or releasable on-chain [#2-claimable-or-releasable-on-chain]
The contract calculates an amount the beneficiary may release. A schedule may vest continuously while claims happen irregularly.
OpenZeppelin’s VestingWallet illustrates this distinction with separate concepts for vested, released, and releasable amounts. It also notes that ownership of an ownable vesting wallet may be transferable, making economic control more complex than a static beneficiary label ([OpenZeppelin](https://docs.openzeppelin.com/contracts-stylus/vesting-wallet)).
### 3. Transferable [#3-transferable]
The beneficiary can move the tokens. Transferability can begin at vesting, after claiming, or after another restriction expires.
### 4. Classified as circulating [#4-classified-as-circulating]
A market-data provider may add some or all unlocked tokens to its circulating-supply estimate according to its methodology. That classification can lag the on-chain event or exclude certain controlled allocations.
### 5. Transferred to a market venue or counterparty [#5-transferred-to-a-market-venue-or-counterparty]
On-chain data may show a transfer to a labeled exchange, market maker, OTC counterparty, DeFi pool, custodian, or another wallet. This is evidence of movement, not definitive evidence of a sale. Exchange labels can also be incomplete or wrong.
The chain of inference should therefore be:
```text
Scheduled → vested → claimable → claimed → transferable → moved → possibly sold
```
Skipping steps turns a public schedule into a false claim about holder intent.
## How do cliff, linear, and step vesting work? [#how-do-cliff-linear-and-step-vesting-work]
### Cliff vesting [#cliff-vesting]
Nothing vests before a specified time; a tranche or the full allocation vests at the cliff.
```text
Vested(t) = 0, for t < cliff
Vested(t) = allocation or scheduled tranche, for t ≥ cliff
```
Check whether vesting accrues during the cliff and becomes available at once, or begins only after the cliff. Both patterns exist.
### Linear vesting [#linear-vesting]
For allocation `A`, vesting start `t0`, and duration `D`:
```text
Vested(t) = 0 if t < t0
Vested(t) = A × (t − t0) / D if t0 ≤ t < t0 + D
Vested(t) = A if t ≥ t0 + D
```
Rounding and timestamp units matter in code. A front end may display monthly unlocks while the contract accrues every second.
### Cliff plus linear vesting [#cliff-plus-linear-vesting]
A common implementation returns zero before the cliff, then applies a linear schedule. Confirm whether the amount accrued since `t0` becomes available at the cliff or the line begins there.
### Step or graded vesting [#step-or-graded-vesting]
Tokens vest in discrete tranches:
```text
Vested(t) = allocation per step × completed steps
```
The schedule might be monthly, quarterly, milestone-based, or irregular.
### Emissions are not always vesting [#emissions-are-not-always-vesting]
Validator or liquidity incentives can be created and distributed according to protocol activity rather than a beneficiary’s locked allocation. Include them in net supply analysis, but label them separately.
## The token unlock verification protocol [#the-token-unlock-verification-protocol]
### Step 1: locate the canonical schedule [#step-1-locate-the-canonical-schedule]
Use this evidence order:
1. Verified vesting or timelock contract
2. Current official tokenomics or transparency disclosure
3. Executed governance proposal
4. Signed allocation agreement, if public
5. Third-party calendar that links to a primary source
Archive the source and record its publication or revision date. Historical schedules can be superseded.
The Solana Foundation’s historical transparency report, for example, separated founders, investors, validators, employees, community allocations, and market-making arrangements, with different timing and restrictions ([Solana Foundation](https://solana.com/news/solana-foundation-transparency-reports)). The lesson is not to reuse those old figures today; it is to preserve recipient-specific detail rather than quote one aggregate unlock number.
### Step 2: reconcile schedule and contract [#step-2-reconcile-schedule-and-contract]
Capture:
* Chain and token contract address
* Vesting contract address
* Beneficiary or beneficiary class
* Total allocation deposited
* Start, cliff, duration, and cadence
* Released amount
* Releasable amount at the research timestamp
* Owner, admin, and upgradeability
* Revocation, acceleration, or transfer rights
If the paper says monthly vesting but the contract is a simple timelock, write the discrepancy. Do not assume the prose overrides deployed logic.
### Step 3: normalize into time buckets [#step-3-normalize-into-time-buckets]
For each recipient class, calculate expected gross unlocks in 30-, 90-, 180-, and 365-day windows.
```text
Gross scheduled unlocks = sum of all tranches in the window
Net scheduled supply change = unlocks + emissions − burns
```
Keep contract-verified values separate from estimates.
### Step 4: measure the event with three denominators [#step-4-measure-the-event-with-three-denominators]
```text
Unlock / current circulation = tokens unlocked / circulating supply
Unlock / current unlocked supply = tokens unlocked / unlocked supply
Unlock notional = tokens unlocked × reference price
```
The notional figure is a scenario at the reference price, not predicted sale proceeds. The first ratio is often more interpretable because it shows how large the tranche is relative to the float the market already sees.
### Step 5: compare with liquidity and recipient behavior [#step-5-compare-with-liquidity-and-recipient-behavior]
Measure:
* Order-book depth at standard price distances
* DEX price impact at relevant sizes
* Median credible spot volume, with methodology
* Recipient transfers after previous releases
* Known delegation, staking, treasury, market-making, or grant use
Use [Crypto Liquidity Analysis](/blog/crypto-liquidity-analysis) for the market-capacity side and [Whale Wallets and Exchange Flows](/blog/whale-wallets-exchange-flows) for label and attribution limits.
## The unlock-impact worksheet [#the-unlock-impact-worksheet]
| Field | Value | Confidence | Primary source |
| -------------------------- | ----- | ---------- | -------------- |
| Current circulating supply | | | |
| Current unlocked supply | | | |
| Next event date/timezone | | | |
| Gross tokens scheduled | | | |
| Beneficiary class | | | |
| Contract / wallet | | | |
| Claim required? | | | |
| Transferable immediately? | | | |
| % of current circulation | | | |
| Reference-price notional | | | |
| Depth / price-impact tests | | | |
| Prior recipient behavior | | | |
| Schedule-change authority | | | |
Add three scenarios:
| Scenario | Assumption | Evidence required |
| ----------------------- | ----------------------------------------- | ------------------------------------------------------------ |
| Low circulation impact | Tokens remain held, staked, or delegated | Post-unlock balances and contract state |
| Partial circulation | Some transfer to venues or counterparties | Verified labeled flows and supply update |
| High circulation impact | Large share becomes market-available | Claims, transfers, provider classification, liquidity change |
Do not assign probabilities unless you have a defensible model and enough comparable observations.
## Worked hypothetical: the “Orchid” cliff [#worked-hypothetical-the-orchid-cliff]
**This example is fictional and not a forecast, recommendation, or observation about any real token.**
Orchid has:
* 250 million circulating tokens
* A maximum supply of 1 billion
* A 40 million investor-token cliff in 30 days
* A reference price of $1.50
* Reported median daily spot volume of $18 million
Headline calculations:
```text
Unlock / circulation = 40m / 250m = 16%
Reference-price notional = 40m × $1.50 = $60m
```
It would be wrong to conclude “$60 million of selling is coming.” Verification finds:
* The contract vests 40 million at the cliff but requires individual claims.
* The vesting-wallet owners can transfer ownership before the cliff.
* Ten million tokens belong to a strategic partner with a public staking commitment, but the commitment is not enforced by the vesting contract.
* Previous quarterly claims ranged from 20% to 65% of the available tranche during the first week.
* A $500,000 simulated sell quote produces materially different impact across the two largest venues.
The responsible output is a range:
> A tranche equal to 16% of current reported circulation becomes claimable. The immediate market-available amount is uncertain because claims are discretionary and one staking commitment is not contract-enforced. Monitor claims, recipient wallets, provider circulation updates, and standardized liquidity tests; do not equate the $60 million notional with expected sale volume.
That conclusion distinguishes known mechanics from unknown behavior.
## How to read unlock behavior after the event [#how-to-read-unlock-behavior-after-the-event]
Check at consistent intervals: event time, +24 hours, +7 days, and +30 days.
Record:
* Amount newly releasable
* Amount claimed or released
* Destination wallets
* Transfers to known contracts, staking, governance, bridges, custodians, or exchange clusters
* Changes in provider-reported circulating supply
* Changes in depth, spread, and price impact
* Relevant market-wide movement
Avoid post hoc causality. Price can move around an unlock because of the broader market, anticipation before the event, hedging, unrelated news, or liquidity conditions. “Price fell after unlock” does not prove unlocked recipients sold.
## Token unlock red flags [#token-unlock-red-flags]
* The schedule has no primary source
* Percentages do not reconcile with the stated supply
* Dates appear only in an image with no revision history
* “Locked” tokens sit in a normal transferable wallet
* Beneficiary or vesting-contract ownership can be transferred but is described as non-transferable
* Admin can accelerate, revoke, replace, or upgrade the schedule without a delay
* The calendar omits emissions, treasury distributions, or market-maker allocations
* Unlocks are quoted against maximum supply to make a near-term event look smaller
* Third-party calendars disagree and none links to current contracts
* A provider updates circulation without explaining the classification
## How unlocks connect to market cap and FDV [#how-unlocks-connect-to-market-cap-and-fdv]
Unlocks are the time dimension missing from the [Market Cap vs FDV](/blog/market-cap-vs-fdv-crypto) ratio.
Two assets can have the same circulation ratio but different risk profiles:
* Asset A releases the gap over 20 years to validators.
* Asset B releases half of the gap to early investors next quarter.
FDV sees the same denominator problem. The unlock schedule reveals timing and recipient concentration. [Tokenomics research](/blog/how-to-research-crypto-tokenomics) adds utility, emissions, and demand.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Legal restrictions or side letters may not be visible on-chain.
* Beneficiaries can use custodians or OTC counterparties that obscure final disposition.
* Entity labels and circulating-supply classifications can be incomplete.
* Wallet transfers do not reveal intent or prove a sale.
* Prior behavior is not a guarantee of future behavior.
* Governance can amend schedules.
* Price and liquidity can change before the event, making old notional comparisons stale.
The correct response is timestamped uncertainty, not invented precision.
## Monitoring an unlock with BlockMind [#monitoring-an-unlock-with-blockmind]
A BlockMind agent can research current disclosures, inspect public contract and wallet pages through its [browser capability](/docs/capabilities/browser), use [on-chain intelligence](/docs/capabilities/on-chain) to review flows and holders, and save the event assumptions in your Notebook. You can ask it to monitor a named condition, while keeping the interpretation under human control.
Example request:
> “Verify the next \[token] unlock from primary sources. Separate scheduled, releasable, claimed, transferable, and provider-circulating amounts. Calculate its share of current circulation, test current liquidity at three sizes, identify every assumption, and set a review for the event plus 24 hours and seven days.”
Always [verify AI-generated crypto analysis](/blog/how-to-verify-ai-crypto-analysis) against the decisive sources.
## The Bottom Line [#the-bottom-line]
An unlock is a change in rights or contract state, not automatic selling. Verify the canonical schedule and live contract, normalize events into time buckets, compare them with current circulation and liquidity, identify recipient incentives, and observe what actually happens after release.
The most honest unlock analysis is a chain of evidence with explicit gaps: scheduled, vested, claimable, claimed, transferable, moved, and only then possibly sold.
## Sources [#sources]
* [Avalanche Builder Hub — Vesting Schedules](https://docs.avax.network/academy/avalanche-l1/l1-native-tokenomics/07-token-distribution/02-vesting-schedules)
* [OpenZeppelin — VestingWallet](https://docs.openzeppelin.com/contracts-stylus/vesting-wallet)
* [OpenZeppelin — Cairo Finance and Vesting Schedules](https://docs.openzeppelin.com/contracts-cairo/3.x/finance)
* [Solana Foundation — Historical Transparency Reports and Unlock Schedule](https://solana.com/news/solana-foundation-transparency-reports)
* [CoinMarketCap — Market Capitalization Methodology](https://support.coinmarketcap.com/hc/en-us/articles/360043836811-Market-Capitalization-Cryptoasset-Aggregate)
---
# Whale Wallets and Exchange Flows: A Verification-First Guide
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: whale wallets and exchange flows
Keywords: whale wallets exchange flows, crypto exchange inflows outflows, whale wallet tracking, on-chain exchange flows, crypto wallet analysis
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** A large wallet transfer proves that assets moved between addresses. It does not, by itself, prove who controls the addresses, why they moved, or whether a trade occurred. Analyze whale and exchange flows through an evidence ladder: verify the transaction, assess label confidence, remove internal and contract mechanics, classify the economic destination, compare with the entity’s history, and require market confirmation before making a stronger interpretation.
***
## What are whale wallets and exchange flows? [#what-are-whale-wallets-and-exchange-flows]
A **whale wallet** is an informal label for an address or entity controlling a large amount of an asset relative to a chosen reference group. There is no universal threshold. Define it—for example, top 0.1% of non-exchange holders or balances above a fixed percentage of circulating supply.
**Exchange inflow** usually means assets transferred into addresses attributed to an exchange. **Exchange outflow** means assets transferred from attributed exchange addresses. **Netflow** is commonly:
```text
Exchange netflow = inflow volume − outflow volume
```
Positive netflow means net movement into the labeled exchange set under that provider’s definitions. It does not automatically mean net selling.
Glassnode’s transparency notice explains that exchange metrics depend on continuously identifying and monitoring exchange-owned addresses and warns that preliminary large flows should be approached with caution ([Glassnode](https://docs.glassnode.com/further-information/exchange-data-transparency-notice)). That dependency is the central limitation: the blockchain records addresses and transfers, while “exchange,” “fund,” and “whale” are analytical labels.
**Scope and assumptions:** This guide addresses public on-chain analysis. It cannot observe an exchange’s internal ledger, private keys, beneficial ownership, OTC settlement, or off-chain intent. It is not a forensic attribution standard or a trade signal.
*This is research, not financial advice. BlockMind’s agent cannot trade, withdraw, transfer, or touch funds.*
## The transfer-to-intent evidence ladder [#the-transfer-to-intent-evidence-ladder]
Use seven levels and do not jump over them:
1. **Transaction fact:** A specific transfer settled on a named chain.
2. **Address label:** A source attributes one or both addresses.
3. **Entity cluster:** Several addresses are grouped under one controller.
4. **Economic transfer:** The move is not merely internal reorganization or contract mechanics.
5. **Destination classification:** Exchange deposit, custody, staking, bridge, DEX, treasury, market maker, or unknown.
6. **Behavior pattern:** The entity’s balances and actions show accumulation, distribution, collateral use, or another repeatable behavior.
7. **Market confirmation:** Price, depth, volume, or executed on-chain trades support the interpretation.
Phrase conclusions at the highest verified level. If you verified only levels 1 and 2, say “transferred to an address labeled as Exchange X,” not “the whale sold.”
## Step 1: verify the raw transaction [#step-1-verify-the-raw-transaction]
Record:
* Chain
* Transaction hash
* Block number and UTC timestamp
* Token contract, not ticker alone
* Amount in native units
* Sender and recipient
* Transfer type: native, token event, internal call, bridge message, mint, burn
* Transaction status
* Related transactions in the same operation
Watch for wrappers, rebasing assets, decimal errors, and tokens that reuse symbols. A bridge deposit on one chain may correspond to a mint on another; counting both as independent economic flows double-counts movement.
## Step 2: grade label confidence [#step-2-grade-label-confidence]
Create an explicit label hierarchy:
| Grade | Evidence | Example conclusion |
| ----- | ------------------------------------------------------------------ | --------------------------------------- |
| A | Entity’s own published address or verifiable signed claim | “Officially disclosed treasury address” |
| B | Multiple reputable providers agree with direct supporting evidence | “High-confidence exchange attribution” |
| C | One provider’s researched label or deterministic cluster | “Provider-labeled exchange address” |
| D | Behavioral heuristic or social claim | “Suspected market-maker wallet” |
| U | No adequate attribution | “Unknown address” |
Nansen documents its human-readable address labels as proprietary data and exposes address- and entity-level analysis through its API ([Nansen API](https://docs.nansen.ai/)). Chainalysis describes a different model that begins with ground-truth attribution and extends clusters through deterministic heuristics ([Chainalysis](https://www.chainalysis.com/blog/chainalysis-data-accuracy/)). Providers can differ in coverage, standards, and update timing. Preserve the provider and retrieval date.
## Step 3: remove non-economic and internal flows [#step-3-remove-non-economic-and-internal-flows]
Before interpreting a large transfer, test these alternatives:
* Exchange hot-to-cold or cold-to-hot wallet movement
* Custodian wallet rotation
* Consolidation of deposit addresses
* UTXO change address
* Bridge lock-and-mint or burn-and-release
* Token migration or redenomination
* Staking deposit or withdrawal
* Collateral movement
* Smart contract upgrade or treasury reorganization
* Market-maker inventory rebalance
* Wrapped-token mint or redemption
Chainalysis explains that centralized exchanges often consolidate many customer deposit addresses into common wallets and that new address evidence can update historical clusters. That means a provider’s historical flow series can change as attribution improves.
For UTXO chains, inspect inputs and change. For account-based chains, decode calls, logs, and protocol contracts. Follow one hop beyond the headline recipient when appropriate, while avoiding speculative address chaining.
## Step 4: classify the destination [#step-4-classify-the-destination]
Use a mutually exclusive primary class plus notes:
| Class | What the transfer supports | What it does not prove |
| ------------------- | ----------------------------------------------------- | ------------------------------------------- |
| Exchange deposit | Assets entered an attributed exchange-controlled path | Sale, order, or final beneficial owner |
| Exchange withdrawal | Assets left attributed exchange control | Long-term holding or accumulation motive |
| DEX swap | An on-chain trade executed under decoded calls | Identity or future intent |
| Liquidity provision | Assets entered a pool/position | Permanent liquidity or bullish intent |
| Staking/delegation | Assets entered a staking path | Price view or lock duration without details |
| Bridge | Assets moved across domains | Net acquisition or disposal |
| Custody | Assets moved to a custodian | Exchange availability or sale |
| Treasury | Official controlled wallet | How funds will be used |
| Unknown | Transfer occurred | Ownership and intent |
DEX swaps can provide stronger evidence of an executed trade than exchange deposits because the swap and amounts are visible. Still account for routing, MEV, arbitrage, and multi-step transactions.
## Step 5: calculate flows with transparent denominators [#step-5-calculate-flows-with-transparent-denominators]
### Gross and net exchange flow [#gross-and-net-exchange-flow]
```text
Gross flow = inflows + outflows
Netflow = inflows − outflows
```
High gross flow with netflow near zero can indicate heavy two-way movement. Netflow alone hides that activity.
### Flow relative to exchange balance [#flow-relative-to-exchange-balance]
```text
Netflow ratio = netflow / beginning labeled exchange balance
```
### Flow relative to circulating supply [#flow-relative-to-circulating-supply]
```text
Supply flow ratio = netflow / circulating supply
```
### Whale balance change [#whale-balance-change]
For a stable defined cohort:
```text
Whale cohort change = ending balances − beginning balances
```
Do not let addresses enter or leave the cohort merely because their balance crossed the threshold without disclosing the composition effect. Entity-adjusted cohorts are preferable when available.
Nansen’s netflow documentation combines DEX activity and CEX transfers in some “smart money” views ([Nansen](https://docs.nansen.ai/api/smart-money/netflows)). That can be useful, but it means you must read the metric definition before comparing it with a provider that reports only exchange-address transfers.
## Step 6: compare with the wallet’s own history [#step-6-compare-with-the-wallets-own-history]
One large transfer can be routine for an entity.
Compare:
* Amount versus its 30-, 90-, and 365-day transfer distribution
* Destination types used previously
* Holding-period changes
* Known vesting, treasury, staking, or market-making schedule
* Transfers around past protocol events
* Whether the address is an active intermediary rather than final holder
Use median and percentile context instead of labeling every large absolute transfer “unusual.”
## Step 7: require market and protocol confirmation [#step-7-require-market-and-protocol-confirmation]
If the interpretation is “potential distribution,” look for:
* Increase in exchange-attributed balance after internal-flow adjustment
* Actual DEX sales or exchange-order evidence, if available
* Deterioration in bid depth or increased price impact
* Spot volume across multiple credible venues
* Recipient wallet balance reduction that cannot be explained by custody or bridging
If the interpretation is “potential accumulation,” look for sustained entity balance growth, exchange withdrawals that reach controlled self-custody addresses, DEX purchases, and absence of immediate recycling.
Use the [Crypto Liquidity Analysis](/blog/crypto-liquidity-analysis) protocol and the broader [Crypto Market Analysis Guide](/blog/crypto-market-analysis-guide). No single flow should override contradictory market evidence.
## Worked hypothetical: the “200 million token exchange inflow” [#worked-hypothetical-the-200-million-token-exchange-inflow]
**This is fictional. It is not a real transfer, market claim, or recommendation.**
An alert says:
> “Whale sends 200 million NOVA to Exchange Z. Dump incoming.”
Verification finds:
1. The transaction is real and equals 4% of NOVA’s reported circulation.
2. The sender has a Grade C label as an Exchange Z cold wallet, not an independent whale.
3. The recipient is a newly activated Exchange Z hot-wallet cluster.
4. Ninety minutes later, many small withdrawals leave the hot wallet.
5. Exchange Z announced wallet maintenance that day.
6. Adjusted entity-level Exchange Z balance changes by only 0.1%.
7. Cross-venue depth and volume show no unusual change.
Responsible conclusion:
> The 200 million NOVA transfer appears to be an internal exchange wallet rotation. Entity-adjusted exchange balance changed minimally, and no market confirmation supports a sale interpretation.
The same raw transaction can support the opposite narrative if labels are ignored.
## A whale-flow verification template [#a-whale-flow-verification-template]
```text
CLAIM
- Exact claim being evaluated:
- Source of alert:
RAW TRANSACTION
- Chain / token contract:
- Hash / block / UTC time:
- Amount and % circulating supply:
- Decoded action:
ATTRIBUTION
- Sender label, provider, grade:
- Recipient label, provider, grade:
- Entity cluster vs single address:
- Conflicting labels:
ALTERNATIVE EXPLANATIONS
- Internal exchange/custody move?
- Bridge/wrapper/migration?
- Staking/collateral/liquidity?
- Market-maker or treasury operation?
BEHAVIOR
- Historical percentile:
- Balance before/after:
- One-hop destination:
- Related schedule/event:
MARKET CONFIRMATION
- DEX trades:
- Exchange balance change:
- Spot volume/depth/impact:
CONCLUSION
- Highest verified evidence-ladder level:
- What remains unknown:
- Next observation trigger:
```
## Common whale and exchange-flow mistakes [#common-whale-and-exchange-flow-mistakes]
### “Exchange inflow means sale” [#exchange-inflow-means-sale]
It means movement into an attributed exchange path. Custody, collateral, internal operations, and later withdrawal remain possible.
### “Exchange outflow means accumulation” [#exchange-outflow-means-accumulation]
It may represent custody rotation, institutional settlement, staking, collateral, or another exchange address not yet labeled.
### “One address equals one person” [#one-address-equals-one-person]
Exchanges, funds, protocols, and individuals can control many addresses; many users can share an exchange cluster.
### “Top holder equals whale” [#top-holder-equals-whale]
The address may be a burn address, bridge, staking contract, pool, treasury, custodian, or exchange.
### “Smart money is always right” [#smart-money-is-always-right]
Labels are based on provider criteria and historical data. Strategies, horizons, and risk constraints differ; past performance classifications can change.
### “On-chain data shows motive” [#on-chain-data-shows-motive]
It shows state transitions. Motive is an inference requiring more evidence.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Labels can be incomplete, stale, or disputed.
* Exchanges create new deposit addresses and move wallets.
* Internal exchange trades are not visible on-chain.
* OTC settlement can look like a transfer without visible price formation.
* Cross-chain bridges and wrappers complicate supply and flow counting.
* Privacy tools and custodians limit entity attribution.
* Historical provider data can be revised after cluster updates.
* Market confirmation can still be coincidental rather than causal.
Always retain raw hashes so another researcher can reproduce the base facts even if labels change.
## Using BlockMind for on-chain flow research [#using-blockmind-for-on-chain-flow-research]
BlockMind’s [on-chain intelligence](/docs/capabilities/on-chain) can help profile public wallets, inspect holder structure, and review whale, smart-money, and exchange flows. Your agent can combine that with current web research and save a verification note in your Notebook.
Ask it to constrain the claim:
> “Verify this whale alert from the raw transaction. Grade every address label, remove internal exchange and bridge mechanics, calculate entity-adjusted balances, compare with the wallet’s history, and separate observed facts from intent hypotheses. Do not call a transfer a buy or sale without execution evidence.”
Then open the decisive transactions and provider definitions yourself. See [How to Verify AI Crypto Analysis](/blog/how-to-verify-ai-crypto-analysis).
## The Bottom Line [#the-bottom-line]
Whale tracking is useful when it remains close to the ledger. Verify the transaction, grade attribution, remove internal mechanics, classify the destination, compare with history, and require market confirmation.
The most trustworthy sentence is often narrower than the alert: “Assets moved to a provider-labeled exchange address; sale intent remains unverified.” Precision about uncertainty is the edge.
## Sources [#sources]
* [Glassnode — Exchange Data Transparency Notice](https://docs.glassnode.com/further-information/exchange-data-transparency-notice)
* [Glassnode — Metric Catalog](https://docs.glassnode.com/data/metric-catalog)
* [Nansen API — Address Labels and Entity Analytics](https://docs.nansen.ai/)
* [Nansen API — Smart Money Netflows](https://docs.nansen.ai/api/smart-money/netflows)
* [Chainalysis — Address Attribution and Deterministic Clustering](https://www.chainalysis.com/blog/chainalysis-data-accuracy/)
* [Etherscan — Address Metadata](https://docs.etherscan.io/metadata/introduction)
---
# What Is an AI Crypto Agent? What a Personal Investing Agent Does All Day
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: informational
Primary keyword: what is an AI crypto agent
Keywords: AI crypto agent, AI investing agent, crypto research agent
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** An AI crypto agent is software that can research, remember context, use market tools, and run scheduled work on your behalf. The useful kind does more than answer prompts: it keeps watch, writes down what it learns, and reports back. It should not control your money. A trustworthy investing agent stays read-only and leaves every trade to you.
***
Most AI investing products are still chatbots with a new label. You ask a question, they answer, and nothing happens until you return.
An actual AI crypto agent has a longer-lived job. It can keep a workspace, remember your portfolio and research, use specialized data sources, wake on a schedule, and tell you when something changed. The difference is not how conversational it sounds. The difference is whether useful work continues between conversations.
## What is an AI crypto agent? [#what-is-an-ai-crypto-agent]
An AI crypto agent is a software analyst that can take a goal, use tools to gather evidence, keep relevant context, and carry out approved tasks without needing a new prompt for every step.
For an investor, that might mean:
* Reading market, on-chain, derivatives, news, and social data
* Connecting findings to the assets you actually hold
* Researching a token or stock into a structured report
* Monitoring a thesis after the first analysis is finished
* Sending a daily briefing and material alerts
* Remembering your preferences, open questions, and prior decisions
The agent is not the source of truth. It is the worker that collects, compares, and explains the sources.
## AI agent, chatbot, or trading bot? [#ai-agent-chatbot-or-trading-bot]
These products solve different problems.
| Product | What it does | What happens when you leave |
| ---------------- | ---------------------------------------------------------- | -------------------------------------------------- |
| General chatbot | Answers the prompt in front of it | Usually waits for the next prompt |
| AI research tool | Produces a report from a question or asset | The report remains, but the research usually stops |
| AI crypto agent | Researches, remembers, monitors, and reports on a schedule | Keeps doing approved work |
| Trading bot | Executes predefined or model-driven orders | May move money automatically |
The last row is the crucial boundary. A personal investing agent does not need trading permission to be useful. In fact, read-only access makes its role clearer: gather the evidence, explain the tradeoffs, and leave execution to the owner.
## What does a personal investing agent do all day? [#what-does-a-personal-investing-agent-do-all-day]
A useful agent's day has four rhythms.
### Morning: prepare the brief [#morning-prepare-the-brief]
Before you open a chart, the agent reviews overnight market moves, relevant news, portfolio changes, and anything you asked it to track. It turns that into a short Morning Brief focused on your holdings, not a generic market recap.
### During the day: research on demand [#during-the-day-research-on-demand]
You can hand it a token, stock, theme, wallet, or question. The agent gathers evidence, checks contradictory signals, and writes a structured analysis. You should be able to see the reasoning and sources, not just a confident conclusion.
### Between visits: monitor what matters [#between-visits-monitor-what-matters]
Research decays. Token unlocks approach, holder concentration changes, funding becomes crowded, a team misses a milestone, or a thesis loses its catalyst. Scheduled monitoring lets the agent compare new evidence with the reason you cared in the first place.
### Over time: keep the notebook [#over-time-keep-the-notebook]
The agent records durable context: research notes, followed ideas, preferences, recurring checks, and the criteria you use to judge ideas. That memory prevents every conversation from starting at zero and makes later updates more useful.
## What should an AI crypto agent know? [#what-should-an-ai-crypto-agent-know]
The best context is specific enough to improve the answer and limited enough to remain safe.
### Your portfolio [#your-portfolio]
If you connect wallets and exchanges, the agent can answer questions against your real exposure. "What does this news mean?" becomes "What does this news mean for the 18% of my portfolio exposed to this sector?"
### Your watchlists and open theses [#your-watchlists-and-open-theses]
A watchlist says what deserves attention. A thesis says why. Monitoring only becomes intelligent when the agent can compare current evidence with an expected catalyst, invalidation level, or risk.
### Your decision process [#your-decision-process]
Two investors can interpret the same setup differently. One may care most about macro liquidity; another may refuse concentrated token ownership. Reusable review criteria let the agent organize evidence around the way you actually judge an idea.
### What it should forget or never receive [#what-it-should-forget-or-never-receive]
An investing agent does not need seed phrases, private keys, withdrawal permissions, or trading authority. It should not ask for them. Sensitive access that does not improve research is risk without benefit.
## What can an AI crypto agent automate well? [#what-can-an-ai-crypto-agent-automate-well]
Agents are strongest at work that rewards breadth, repetition, and consistency:
* Gathering the same categories of evidence for every asset
* Comparing several data sources for contradictions
* Summarizing long documents and news flows
* Monitoring a known condition on a schedule
* Re-running a checklist without skipping the boring steps
* Preparing a first draft of a research conclusion
This is the data-heavy part of due diligence. Our [AI vs manual crypto research guide](/blog/ai-vs-manual-crypto-research) maps the split in detail.
## What should stay human? [#what-should-stay-human]
The final judgment should stay yours, especially when the decision depends on:
* Whether a founder or team deserves trust
* Whether an incentive can survive contact with real behavior
* Whether a narrative is early, crowded, or already exhausted
* How a position fits your time horizon and financial life
* How much uncertainty and downside you can afford
An agent can make these questions explicit and present evidence. It cannot own the consequences.
## Can an AI investing agent trade for you? [#can-an-ai-investing-agent-trade-for-you]
Some trading agents and bots can. A research agent does not need to.
BlockMind's boundary is deliberately read-only: wallet and exchange connections let the agent see positions, but it cannot trade, withdraw, or move funds even if you told it to. You remain the only person who can act.
That boundary also makes recommendations easier to evaluate. The product has no execution loop pushing it to turn every observation into a trade.
## How BlockMind's agent works [#how-blockminds-agent-works]
BlockMind gives Pro members one named personal AI investing agent with its own workspace. Its work is organized around four stages:
1. **Explore** scans crypto and stock markets for ideas.
2. **Analyze** turns a saved idea into a structured report.
3. **Verdicts** asks a panel of experts for a call: approved, watch, wait, or reject.
4. **Track** follows holdings, research, and the verdicts you care about.
By default, the agent publishes a Morning Brief to the dashboard and sends it by email at about 8:00 in your timezone. The agent can also run monitoring checks during the day and deliver in-app or email alerts. Its [Notebook](/docs/memory/notebook) keeps research and preferences available for later conversations.
The agent comes with Pro. The free classic platform still covers portfolio tracking, watchlists, charts, and market indicators without running the personal agent.
## How to evaluate an AI crypto agent [#how-to-evaluate-an-ai-crypto-agent]
Before trusting one with your attention, ask:
1. **Does it show sources?** A conclusion without evidence is just synthetic confidence.
2. **Does it know my actual portfolio?** Generic market commentary is not personalization.
3. **Does it work between prompts?** Scheduled monitoring separates an agent from a chat window.
4. **Can I inspect and change its memory?** Persistent context should be visible and correctable.
5. **What permissions does it require?** Research should work with read-only access.
6. **Does it state limits clearly?** Look for honest update frequency, coverage, and plan terms.
7. **Can it explain disagreement?** A useful analyst surfaces uncertainty instead of hiding it.
## Frequently Asked Questions [#frequently-asked-questions]
### Is an AI crypto agent the same as an AI trading bot? [#is-an-ai-crypto-agent-the-same-as-an-ai-trading-bot]
No. A trading bot executes orders. A research agent gathers evidence, monitors conditions, and explains what changed. Some products combine both, so always inspect the permissions. BlockMind stays read-only and cannot trade or move funds.
### Does an AI agent replace crypto research? [#does-an-ai-agent-replace-crypto-research]
It replaces much of the gathering and repetition, not the final judgment. You still need to assess credibility, incentives, timing, and position size.
### Does an agent need my wallet keys? [#does-an-agent-need-my-wallet-keys]
No. A research agent can read public wallet addresses or use provider-scoped read-only exchange keys where available. BlockMind's Firi connection is the exception because Firi does not offer that key scope; BlockMind enforces balance-only behavior in the application. Never provide a seed phrase, private key, or withdrawal permission.
### Can an AI agent monitor my portfolio while I sleep? [#can-an-ai-agent-monitor-my-portfolio-while-i-sleep]
Yes, if the product supports scheduled work. Check how often monitoring runs and where alerts arrive. "Always-on" should not be confused with tick-level, real-time execution.
### How much does a BlockMind agent cost? [#how-much-does-a-blockmind-agent-cost]
See [current BlockMind plans, introductory terms, and the card requirement](/docs/overview/plans-and-pricing) on the canonical pricing page. The classic tracking platform and public DeepDive reports remain available separately for free within their stated limits.
## The Bottom Line [#the-bottom-line]
An AI crypto agent is valuable when it turns research from an occasional burst into a maintained habit. It gathers the evidence, remembers the thesis, checks what changed, and puts the important part in front of you.
The trust model matters as much as the intelligence. Give the agent context, not custody. Let it do the repetitive work, demand sources for every important claim, and keep the final decision in human hands.
## Learn More [#learn-more]
* [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research)
* [Can AI Replace a Crypto Research Analyst?](/blog/can-ai-replace-crypto-research-analyst)
* [Crypto Morning Brief](/blog/crypto-morning-brief)
* [AI Portfolio Monitoring](/blog/ai-portfolio-monitoring-crypto)
* [AI Crypto Analyst vs ChatGPT](/blog/ai-crypto-analyst-vs-chatgpt)
* [The BlockMind research journey](/docs/capabilities/research-journey)
* [BlockMind trust and security](/docs/security)
## Sources [#sources]
1. [NIST: AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework), accessed July 2026.
2. [Coinbase: API key authentication and permissions](https://docs.cdp.coinbase.com/coinbase-app/authentication-authorization/api-key-authentication), accessed July 2026.
3. [CFTC: Customer Advisory on AI Trading Bots](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html), January 2024.
---
# What Is On-Chain Analysis in Crypto? A Practical Framework
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: Market and On-chain Signals
Search intent: informational
Primary keyword: what is on-chain analysis in crypto
Keywords: what is on-chain analysis in crypto, on-chain analysis, crypto on-chain data, blockchain analytics, AI on-chain analysis
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** On-chain analysis is the process of turning public blockchain records into evidence about transactions, balances, ownership patterns, protocol usage, liquidity, and capital flows. The record can prove that an address called a contract at a block; it usually cannot prove who controlled the address, why they acted, or what happened on a centralized exchange. Good analysis keeps raw facts, decoded actions, entity labels, and economic interpretations separate.
“The blockchain is transparent” is true but incomplete. A chain stores machine-readable state and events, not labels such as “whale accumulation,” “real user,” or “protocol revenue.” Those conclusions require indexing, decoding, clustering, methodology, and judgment.
Use this evidence framework inside the wider [crypto market analysis guide](/blog/crypto-market-analysis-guide), which combines on-chain findings with liquidity, sentiment, technical, and regime context.
## The four layers of on-chain evidence [#the-four-layers-of-on-chain-evidence]
### Layer 1: Raw chain records [#layer-1-raw-chain-records]
At the base are blocks, transactions, logs, traces, balances, and contract state. Ethereum’s analytics documentation describes raw tables for blocks, transactions, event logs, and call traces ([Ethereum.org](https://ethereum.org/developers/docs/data-and-analytics/)). These records can answer questions such as:
* Did a transaction occur?
* Which address initiated it?
* Which contract was called?
* What token transfer events were emitted?
* What was an address balance at a stated block?
* How much gas was used?
This is the most reproducible layer when chain, block, transaction hash, and decoding assumptions are stated.
### Layer 2: Decoded actions [#layer-2-decoded-actions]
Contract interfaces and protocol schemas turn bytes and events into actions such as swap, deposit, borrow, repay, bridge, mint, or vote. Decoding can fail when:
* the wrong contract interface is used;
* a proxy implementation changes;
* an aggregator routes one user action through several protocols;
* a protocol records important state without a convenient event;
* a cross-chain action is visible only in pieces.
“Transfer” is an event type, not an economic explanation. A transfer may be a purchase settlement, internal reorganization, bridge lock, staking deposit, collateral move, vesting distribution, or gift.
### Layer 3: Entity and behavior labels [#layer-3-entity-and-behavior-labels]
Analytics providers add labels such as exchange, bridge, protocol treasury, market maker, fund, or smart money. Nansen documents entity and behavioral address labels ([Nansen](https://docs.nansen.ai/api/profiler/address-labels)) and separately defines its “Smart Money” dataset as a curated list ranked using realized profit, win rate, and performance across market cycles ([Nansen](https://docs.nansen.ai/about/endpoints-overview)).
These labels make raw data interpretable, but they are not chain consensus. A label can be incomplete or wrong, an entity can change addresses, and a single exchange wallet can represent many customers.
### Layer 4: Economic interpretation [#layer-4-economic-interpretation]
The analyst connects the observed action to a question:
* Is usage growing or are incentives recycling capital?
* Does a treasury transfer create possible future selling pressure?
* Is exchange inflow broad or dominated by one entity?
* Do fees accrue to the protocol, token holders, validators, or another party?
* Does a holder trend reflect new investors or address splitting?
This layer carries the most decision value and the most uncertainty. It should be presented as inference with alternatives, not as an immutable on-chain fact.
## What on-chain analysis can measure [#what-on-chain-analysis-can-measure]
### Network activity [#network-activity]
* transactions and contract calls;
* active or new addresses under a declared definition;
* fees paid and gas used;
* validator or staking participation;
* settlement and bridge activity.
Address activity is not user activity. One user can control many addresses; one custodian address can represent many users. Glassnode’s entity research explicitly describes this two-way problem and uses clustering to estimate entities rather than equate addresses with people ([Glassnode](https://research.glassnode.com/bitcoin-supply-distribution/)).
### Token ownership and flows [#token-ownership-and-flows]
* holder balances and concentration;
* team, treasury, and vesting movements;
* exchange deposits and withdrawals;
* bridge escrow and wrapped supply;
* mint, burn, and emission events;
* labeled entity accumulation or distribution.
The correct workflow classifies infrastructure before interpreting concentration. See [how to analyze crypto holder distribution](/blog/crypto-holder-distribution).
### Protocol activity [#protocol-activity]
* deposits, borrows, repayments, and withdrawals;
* decentralized-exchange volume and liquidity;
* collateral composition;
* fees and revenue under a stated methodology;
* liquidations;
* governance proposals and votes.
Total value locked can change because users deposit or withdraw, because token prices move, or because the provider changes mappings. Decompose quantity and valuation effects where possible.
### Market structure [#market-structure]
* decentralized-exchange pool depth and swap activity;
* stablecoin flows;
* funding collateral movements when on-chain;
* miner, validator, or treasury selling pressure;
* realized-cap and coin-age metrics for UTXO chains;
* cross-chain liquidity migration.
On-chain data does not reveal the full centralized order book or every beneficial owner. Combine it with market data rather than treating it as a complete market view.
For a focused application of labeled flows and their failure modes, read [how to interpret whale wallets and exchange flows](/blog/whale-wallets-exchange-flows).
## A practical on-chain research workflow [#a-practical-on-chain-research-workflow]
### 1. Define the claim precisely [#1-define-the-claim-precisely]
Replace “whales are buying” with:
> Did high-confidence labeled entities increase net balances of the canonical TOKEN contract on chain C between blocks A and B, excluding bridge, exchange, vesting, and internal transfers?
The narrower question tells you what data and labels are required.
### 2. Resolve identifiers [#2-resolve-identifiers]
Record chain, contract address, token decimals, block range, timezone, and whether proxies or bridged representations exist. Stop when the asset is ambiguous.
### 3. Preserve raw evidence [#3-preserve-raw-evidence]
Keep transaction hashes, block heights, raw balances, and event logs before applying exclusions. This lets another researcher challenge the interpretation without recreating your retrieval.
### 4. Decode and classify [#4-decode-and-classify]
Map events to protocol actions and addresses to high-confidence categories. Assign a label source and confidence. “Unknown” is a legitimate category.
### 5. Normalize denominators and time [#5-normalize-denominators-and-time]
Use consistent supply, USD price timestamp, chain coverage, and window. Do not combine a weekly flow with a monthly balance change as if they measured the same period.
### 6. Test alternative explanations [#6-test-alternative-explanations]
For every result, list at least one alternative:
| Observation | Possible interpretation | Alternative explanation |
| --------------------- | -------------------------- | ---------------------------------------------- |
| Exchange inflow | Potential sale preparation | Custody transfer or exchange wallet reshuffle |
| New addresses rise | Adoption | Airdrop farming or address splitting |
| TVL rises in USD | New deposits | Existing assets appreciated |
| Treasury sends tokens | Distribution/sale risk | Grant, market-making, or internal custody move |
| Whale balance rises | Accumulation | Relabeling, OTC settlement, or bridge movement |
### 7. Cross-check off-chain evidence [#7-cross-check-off-chain-evidence]
Read governance proposals, token unlock disclosures, audit reports, incident reports, and exchange announcements. On-chain evidence can confirm what moved; off-chain documents often explain authorized purpose.
### 8. State the conclusion at the right confidence [#8-state-the-conclusion-at-the-right-confidence]
A strong conclusion looks like:
> The treasury multisig transferred X tokens at block Y to an address labeled by provider Z as exchange-associated. The transfer is verified; beneficial ownership and sale are not. No corresponding swap was observed in the covered on-chain venues. The movement increases potential liquid supply but does not prove disposal.
## How AI makes on-chain analysis easier [#how-ai-makes-on-chain-analysis-easier]
AI can translate plain-language questions into retrieval tasks, join market and web evidence, explain contract actions, compare periods, and maintain a repeatable report format. It is especially useful for:
* resolving what a metric means;
* generating and checking queries;
* classifying findings into fact and inference;
* finding contradictory labels;
* comparing a new event with a saved thesis;
* writing a concise explanation from structured results.
It should not invent a wallet owner, treat a label as certain, or claim motive from one transaction. A useful prompt is:
> Show the raw transactions and block range first. Then list each decoding and entity-label assumption with its source. Separate observed actions from inferred economic meaning, and provide at least two plausible alternative explanations.
BlockMind’s [on-chain intelligence capability](/docs/capabilities/on-chain) can combine holder and wallet analysis with protocol fundamentals and portfolio research in plain language. The same evidentiary limits still apply.
## Five common on-chain metrics people misread [#five-common-on-chain-metrics-people-misread]
### Active addresses [#active-addresses]
Not unique people. Methodologies differ on whether to count senders, receivers, contracts, zero-balance addresses, or same-day repeats.
### Transaction count [#transaction-count]
Not equal economic activity. Cheap chains, bots, routing, failed calls, and spam can inflate count.
### Exchange netflow [#exchange-netflow]
Depends on address labels and internal-wallet filtering. Inflow may indicate potential supply, not an executed sale.
### Holder count [#holder-count]
Can rise through dusting, airdrops, sybil behavior, or splitting. Look at balance distribution and meaningful activity.
### Protocol revenue [#protocol-revenue]
Fees paid by users, fees retained by a protocol, validator payments, liquidity-provider earnings, and token-holder accrual are different concepts. Use the provider’s definition.
## What on-chain analysis cannot tell you alone [#what-on-chain-analysis-cannot-tell-you-alone]
* the legal identity behind every address;
* private agreements or beneficial ownership;
* trades inside a centralized exchange;
* team competence or honesty;
* future demand;
* whether a wallet movement was rational;
* a guaranteed price direction;
* complete activity on systems that store data off-chain.
Ethereum’s documentation notes that rollups can execute transactions off-chain and post compressed batches or proofs to the base layer; data availability and historical retrieval vary by design ([Ethereum.org](https://ethereum.org/developers/docs/data-availability)). “It is not visible on Ethereum L1” does not necessarily mean it did not happen.
## Limitations and counterevidence [#limitations-and-counterevidence]
Public ledgers make many claims unusually auditable. They can expose discrepancies between a project’s story and actual funds. But transparency can create false confidence: raw records are exact while the labels and narratives built on top are uncertain.
Provider methodologies also change. Recalculate historical comparisons under one consistent version where possible, and record retrieval dates. Cross-provider agreement may strengthen confidence, but providers can share upstream data or heuristics.
## The Bottom Line [#the-bottom-line]
On-chain analysis is evidence engineering. Start with a precise claim, preserve raw records, decode actions, add labels with confidence, normalize denominators, test alternatives, and cross-check off-chain sources. The blockchain can tell you what an address did; careful analysis determines what you can responsibly infer.
This is research, not financial advice. BlockMind’s agent never tells you what to buy or sell and cannot touch funds.
## Sources [#sources]
* [Ethereum.org — Data and Analytics](https://ethereum.org/developers/docs/data-and-analytics/)
* [Ethereum.org — JSON-RPC API](https://ethereum.org/developers/docs/apis/json-rpc/)
* [Ethereum.org — Data Availability](https://ethereum.org/developers/docs/data-availability)
* [Nansen API — Address Labels](https://docs.nansen.ai/api/profiler/address-labels)
* [Nansen API — Endpoints Overview](https://docs.nansen.ai/about/endpoints-overview)
* [Glassnode Research — Bitcoin Supply Distribution and Entities](https://research.glassnode.com/bitcoin-supply-distribution/)
---
# What Questions Can You Ask AI About Crypto? 36 Useful Questions
Published: 2026-07-10
Updated: 2026-07-10
Review by: 2026-10-10
Author: BlockMind Research Team
Topic: AI Research and Agents
Search intent: informational
Primary keyword: questions to ask AI about crypto
Keywords: questions to ask AI about crypto, crypto AI prompts, AI crypto research questions, crypto research prompts, how to use AI for crypto research
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** The most useful questions to ask AI about crypto are not price predictions. Ask it to define the asset, retrieve timestamped evidence, test the project and token, map risks to your portfolio, argue against the thesis, and specify what would change the conclusion. Require sources, dates, assumptions, and an explicit list of unknowns in every consequential answer.
You can ask AI to explain a protocol, compare competitors, inspect token distribution, summarize governance, analyze a portfolio file, or monitor a research question. Current general-purpose systems can search the web and cite sources, while data-analysis tools can inspect structured files such as CSVs. Those capabilities make AI useful for research—but they do not make its unsourced claims true. OpenAI itself advises users to verify important outputs because a model can still produce fabricated facts or citations ([OpenAI](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)).
The best prompt is therefore not a clever sentence. It is a small research brief.
This question ladder is the practical companion to [AI vs manual crypto research](/blog/ai-vs-manual-crypto-research), which defines which parts of the work to automate and which judgments to keep.
## Start with an output contract [#start-with-an-output-contract]
Add this block to any high-stakes crypto question:
> Use primary sources where possible. Give each time-sensitive figure a source and timestamp. Distinguish facts, estimates, and inference. State which chain, contract address, trading pair, venue, currency, and time window you used. List missing evidence and the strongest counterargument. Do not give a buy or sell instruction.
That request solves several recurring problems at once:
* **Entity ambiguity:** a ticker can refer to more than one token.
* **Time ambiguity:** “current” can mean the last trade, a cached quote, or a daily close.
* **Denominator mistakes:** total supply, circulating supply, and free float are not interchangeable.
* **Source laundering:** a polished summary can hide a weak or circular source.
* **False certainty:** a conclusion can sound precise even when key evidence is unavailable.
## 36 crypto AI questions, organized by research stage [#36-crypto-ai-questions-organized-by-research-stage]
Use the questions in order when you are evaluating an unfamiliar asset. Later questions depend on evidence gathered earlier.
### 1. Establish identity and scope [#1-establish-identity-and-scope]
1. “Identify the official project, token contract, chain, ticker, and canonical documentation. Flag any same-ticker assets.”
2. “Explain what this protocol does in 150 words, then describe the user, problem, and transaction that create demand.”
3. “Separate the company, protocol, network, and token. Which one captures value, if any?”
4. “List the system’s trusted parties, admin controls, oracles, bridges, and upgrade mechanisms.”
5. “What facts are directly verifiable on-chain, and what important facts remain off-chain?”
6. “Build a glossary of the five concepts I must understand before reading the whitepaper.”
This first stage prevents a common failure: doing deep analysis on the wrong contract or assuming that product usage automatically benefits a token.
### 2. Verify the project and people [#2-verify-the-project-and-people]
7. “Find the official team claims, then independently verify each founder’s prior roles and shipped work.”
8. “Compare the roadmap with dated releases, repositories, governance proposals, and product evidence.”
9. “Show the repository’s contributor pattern, release cadence, and recent substantive changes. Do not treat commit count alone as quality.”
10. “List audits, the exact code or deployment each audit covered, unresolved findings, and changes made after the audit.”
11. “Find material incidents, exploits, outages, abandoned versions, or regulator actions. Include the project’s response.”
12. “What evidence would distinguish a real but early project from a polished marketing shell?”
For a manual version of this investigation, use the [crypto team verification checklist](/blog/how-to-check-crypto-team-legit) and the broader [12-step DYOR checklist](/blog/manual-dyor-checklist-12-steps).
### 3. Analyze token economics and ownership [#3-analyze-token-economics-and-ownership]
13. “Create a supply table: current on-chain supply, reported circulating supply, maximum supply, emissions, burns, and next unlocks.”
14. “Reconcile market capitalization and fully diluted valuation using the same price timestamp.”
15. “Classify the top holders as exchanges, bridges, liquidity pools, treasuries, vesting contracts, burn addresses, or likely independent holders.”
16. “Calculate top-10 and top-20 concentration after excluding clearly identified infrastructure addresses. Show both raw and adjusted figures.”
17. “Trace material transfers from team, treasury, or vesting wallets over the selected period. What is observed, and what is only inferred?”
18. “Explain how the token is used. Which uses create durable demand, and which merely recycle incentives?”
Raw holder rankings are easy to misread because one exchange address can represent many customers and one person can control many addresses. Read [how to check token holder distribution](/blog/crypto-holder-distribution) before treating an address count as an investor count.
### 4. Read market, on-chain, and social evidence [#4-read-market-on-chain-and-social-evidence]
19. “Give the price, venue, pair, timestamp, 24-hour volume, and spread. Explain whether the quote is aggregated or venue-specific.”
20. “Compare spot volume with visible liquidity and estimate price impact for three hypothetical order sizes.”
21. “Describe trend, volatility, and nearby support or resistance as zones—not guaranteed reversal points.”
22. “Compare active addresses, transactions, fees, and protocol revenue over consistent windows. Identify obvious counting artifacts.”
23. “Separate deposits, withdrawals, bridge transfers, internal wallet movements, and economically meaningful activity.”
24. “Measure social attention, tone, source diversity, authenticity, and persistence. Which component changed first?”
Blockchains expose transactions and state, but interpretation usually requires indexing, decoding, and entity labels. Ethereum’s developer documentation describes raw analytics data as blocks, transactions, logs, and traces—not prewritten investment conclusions ([Ethereum.org](https://ethereum.org/developers/docs/data-and-analytics/)). For a fuller framework, see [what on-chain analysis means](/blog/what-is-on-chain-analysis-crypto) and [how to read crypto social sentiment](/blog/how-to-read-crypto-social-sentiment).
### 5. Make the research portfolio-aware [#5-make-the-research-portfolio-aware]
25. “Using this holdings file, calculate asset, chain, venue, stablecoin, and narrative concentration. Show formulas.”
26. “Group economically related exposures—for example, ETH, liquid-staking tokens, and Ethereum beta—without pretending the correlations are fixed.”
27. “Which single event or dependency affects the largest share of this portfolio?”
28. “Stress-test three scenarios and show assumptions rather than assigning fake probabilities.”
29. “Which positions contribute most to historical volatility and drawdown in the supplied period?”
30. “What data is missing for a reliable performance calculation: cost basis, transfers, fees, staking rewards, or derivatives?”
ChatGPT can analyze a properly structured spreadsheet, but its analysis environment cannot fetch external data by itself; OpenAI says you must upload or connect the needed external data ([OpenAI data analysis guide](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt/)). A dedicated, read-only portfolio tool can reduce manual updates. Compare both workflows in [Does ChatGPT know your crypto portfolio?](/blog/chatgpt-crypto-portfolio).
### 6. Challenge the decision [#6-challenge-the-decision]
31. “Write the strongest evidence-based case against this thesis. Do not merely list generic crypto risks.”
32. “Which claim in the thesis depends on the weakest source?”
33. “What base-rate comparison am I ignoring? Compare this project with failed and successful peers at a similar stage.”
34. “Create a decision journal with thesis, evidence, counterevidence, unknowns, invalidation conditions, and review date.”
35. “What observable event would update the conclusion, and where would I verify it?”
36. “Summarize the research as facts, interpretations, scenarios, and unresolved questions. Do not collapse them into a score.”
These questions change AI from an answer generator into a criticism engine. That is usually the higher-value use.
## Weak question, stronger question [#weak-question-stronger-question]
| Weak prompt | Stronger research prompt |
| --------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| “Is this token good?” | “Define five criteria, collect primary evidence for each, give counterevidence, and leave the decision to me.” |
| “Why is BTC up?” | “For the last 24 hours, separate verified events, market data, and plausible interpretations; cite timestamps.” |
| “Are whales buying?” | “Define whale, chain, token, entity labels, window, and whether transfers represent accumulation, exchange deposits, or internal movement.” |
| “What is the next support?” | “Identify zones using two stated methods and a fixed data window; show what would invalidate each zone.” |
| “Rate my portfolio risk.” | “Calculate concentration and scenario exposure from the supplied holdings, then list what cannot be inferred without cost basis and liabilities.” |
## What AI should not decide for you [#what-ai-should-not-decide-for-you]
Do not outsource custody, transaction approval, position sizing, risk tolerance, or the final investment decision. Never paste a seed phrase, private key, backup code, or unrestricted exchange credential into an AI conversation. A legitimate research workflow does not need them.
Regulators are equally direct about prediction claims. The CFTC warns that AI cannot predict sudden market changes and treats guaranteed-return or perfect-win-rate claims as scam signals ([CFTC](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)). Use AI to surface evidence and expose contradictions, not to manufacture certainty.
## Limitations and counterevidence [#limitations-and-counterevidence]
* Search can improve freshness but may retrieve a cached page, weak source, or wrong asset.
* On-chain data is transparent, yet wallet ownership labels are probabilistic and centralized-exchange trades are largely off-chain.
* Sentiment can follow price rather than lead it, and coordinated accounts can distort attention.
* Portfolio analysis is only as complete as the positions, liabilities, prices, and transaction history supplied.
* A cited answer can still misquote or overinterpret a genuine source. Open the source.
## The Bottom Line [#the-bottom-line]
Ask AI questions that produce an auditable research trail: identity, evidence, method, timestamp, counterargument, unknowns, and update conditions. The winning prompt is not “What should I buy?” It is “What do we know, how do we know it, what could make it wrong, and what should I verify next?”
BlockMind can run market, on-chain, social, and portfolio research in a persistent workflow, but it remains research—not financial advice. Your agent never tells you what to buy or sell and cannot touch your funds. See [what your agent can do](/docs/overview/what-your-agent-can-do) for the current product scope.
This article is research, not financial advice.
## Sources [#sources]
* [OpenAI — Does ChatGPT tell the truth?](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth)
* [OpenAI — Data analysis with ChatGPT](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt/)
* [Ethereum.org — Data and analytics](https://ethereum.org/developers/docs/data-and-analytics/)
* [CFTC — AI Won’t Turn Trading Bots into Money Machines](https://www.cftc.gov/LearnAndProtect/AdvisoriesAndArticles/AITradingBots.html)
---
# What Data Should You Check Before Buying Any Cryptocurrency?
Published: 2026-03-10
Updated: 2026-07-10
Review by: 2027-01-10
Author: BlockMind Research Team
Topic: Crypto Due Diligence
Search intent: problem-solving
Primary keyword: what to check before buying crypto
Keywords: crypto buying checklist, crypto due diligence, research cryptocurrency
Editorial policy: https://docs.blockmind.app/docs/editorial-policy
**Key takeaway:** Before buying any cryptocurrency, check the same core data every time: market cap, trading volume, liquidity, holder distribution, team credibility, audit history, tokenomics, social activity, and on-chain usage. If you skip even a few of these, you are not investing. You are guessing.
***
Most crypto losses do not come from picking the "wrong narrative." They come from buying assets you did not properly understand. The core checklist has not changed: you need valuation data, liquidity data, ownership data, project credibility, and on-chain evidence. In its 2026 Crypto Crime Report, Chainalysis estimated that scams and fraud received at least **$14 billion on-chain in 2025** and projected that the figure could exceed **$17 billion** as more illicit addresses are identified. That is why a real crypto buying checklist matters: not to predict winners perfectly, but to reduce avoidable mistakes before you commit capital. ([Chainalysis, January 2026](https://www.chainalysis.com/blog/crypto-scams-2026/))
## What should you check before buying crypto? [#what-should-you-check-before-buying-crypto]
You should check whether the asset is liquid enough to trade, distributed safely enough to avoid obvious dump risk, credible enough to deserve trust, and active enough on-chain to justify attention. That means looking at both market data and project quality.
A simple way to think about it is this:
* **Market data** tells you how the token is priced
* **Liquidity data** tells you whether you can realistically enter and exit
* **Ownership data** tells you who can move the market against you
* **Project data** tells you whether the team and product are real
* **On-chain data** tells you whether anyone is actually using it
* **Sentiment data** tells you whether the move is driven by substance or hype
If you only check price, you are missing most of the picture.
## Why isn't price alone enough when buying a cryptocurrency? [#why-isnt-price-alone-enough-when-buying-a-cryptocurrency]
Price alone is not enough because a cheap token can still be wildly overvalued, illiquid, insider-controlled, or fundamentally broken. A token at $0.02 is not automatically "cheaper" than one at $2,000.
Price without context causes three common mistakes:
1. **Ignoring supply:** A low unit price can still imply a huge valuation if supply is massive.
2. **Ignoring liquidity:** A token can be up 40% on paper but impossible to exit without heavy slippage.
3. **Ignoring quality:** Scams and weak projects can look strong for short periods when speculation is high.
This is the same reason broader market context matters. A token that looks attractive during [extreme greed](/blog/fear-greed-index-guide) may look very different once hype cools off.
## What market cap data should you check before buying crypto? [#what-market-cap-data-should-you-check-before-buying-crypto]
You should check both **market cap** and **fully diluted valuation (FDV)** before buying crypto. Market cap tells you the current value of circulating supply. FDV tells you what the project would be worth if all tokens were already in circulation.
Start with these questions:
* What is the current market cap?
* What is the FDV?
* Is the gap between market cap and FDV small or huge?
* Does the valuation make sense relative to adoption, revenue, or usage?
A useful rule of thumb:
* **Small gap between market cap and FDV** = lower dilution risk from future unlocks
* **Huge gap between market cap and FDV** = future supply may pressure price even if the project looks popular today
This matters most in newer tokens. A project can look strong while most of the supply is still locked for insiders, treasury, or future incentives. For a fast market-cap sanity check, compare the token against sector peers and broader market data on sources like [CoinGecko global charts](https://www.coingecko.com/en/global-charts).
## What volume should you check before buying a coin? [#what-volume-should-you-check-before-buying-a-coin]
You should check whether trading volume is high enough, consistent enough, and real enough to support your position size. Volume shows how actively a token trades. Without volume, the price can move fast in both directions.
Look for three things:
* **24h trading volume** relative to market cap
* **Consistency** across multiple days, not just one spike
* **Quality of trading venues** where volume is happening
Red flags include:
* Volume that suddenly appears after long inactivity
* Most volume concentrated on obscure venues
* Very high price volatility with thin order books
* A token that trends on social media but still has weak real trading depth
High volume does not guarantee safety, but low-quality volume often signals fragility.
## How do you check liquidity before buying a cryptocurrency? [#how-do-you-check-liquidity-before-buying-a-cryptocurrency]
You check liquidity by asking how easy it is to buy or sell without moving the price too much. Liquidity is one of the most under-checked data points in crypto, even though it directly affects execution risk.
Check:
* How deep the liquidity pools or order books are
* Whether liquidity is spread across credible venues or concentrated in one place
* Whether buy and sell slippage becomes severe on modest order sizes
* Whether liquidity can be removed quickly by a small set of actors
For decentralized tokens, liquidity matters as much as price. A token can show a strong chart while sitting on fragile liquidity. That is one reason [rug pulls](/blog/5-signs-crypto-rug-pull) happen so fast: once liquidity disappears, the quoted price stops meaning much.
If you are buying small-cap tokens, this check is not optional.
## Why should you check holder distribution before buying crypto? [#why-should-you-check-holder-distribution-before-buying-crypto]
You should check holder distribution because concentrated ownership creates dump risk, governance risk, and price manipulation risk. If a few wallets control too much of the supply, they control your downside more than the chart does.
Look for:
* The percentage held by the top 10 or top 20 wallets
* Whether the largest wallets belong to exchanges, team wallets, treasury wallets, or unknown addresses
* Whether insider allocations are locked or liquid
* Whether whale wallets are accumulating, distributing, or staying flat
Holder concentration is one of the fastest ways to detect hidden fragility. A project can market itself as "community-owned" while a small group still controls the outcome.
This is also where on-chain context matters. Wallet concentration on its own is not enough; you also want to know whether those large holders are moving coins to sell, holding steady, or absorbing supply.
## What should you check about the team and project credibility? [#what-should-you-check-about-the-team-and-project-credibility]
You should check whether the team is identifiable, accountable, and capable of shipping. Crypto moves fast, but credibility still matters.
Start with the basics:
* Are founders and core contributors public?
* Do they have a real track record?
* Is the documentation clear?
* Is there an active product, protocol, or user base?
* Are recent updates substantive or just marketing?
Healthy signs:
* Transparent leadership or at least a well-established reputation
* Clear product positioning
* Evidence of ongoing development
* A believable roadmap tied to actual execution
Red flags:
* Anonymous team with no long-term reputation
* Vague promises with no proof of delivery
* Dead GitHub, dead docs, or dead socials
* Heavy influencer marketing with weak fundamentals
Not every anonymous team is fraudulent, but anonymity raises the burden of proof.
## Why do audits matter before buying a token? [#why-do-audits-matter-before-buying-a-token]
Audits matter because unaudited or poorly reviewed smart contracts increase the chance of exploit risk, hidden permissions, and catastrophic loss. An audit is not a guarantee, but no audit is a major warning sign for any serious on-chain project.
Check:
* Whether the project has been audited
* Which firm performed the audit
* Whether critical findings were fixed
* Whether the deployed contracts match what was reviewed
* Whether admin privileges, upgrade rights, or pause functions remain highly centralized
What audits do well:
* Catch obvious logic flaws
* Surface risky permissions
* Expose missing safeguards
What audits do **not** do:
* Guarantee safety
* Protect against bad token economics
* Protect against insider dumping
* Eliminate governance risk
Security should be treated as one layer in the checklist, not the whole checklist.
## What tokenomics should you check before buying crypto? [#what-tokenomics-should-you-check-before-buying-crypto]
You should check supply structure, unlock schedule, insider allocation, and incentive design before buying crypto. Tokenomics determines how value can leak out of the system even when the story sounds good.
Review these items:
* Total supply and circulating supply
* Vesting schedule for team and investors
* Upcoming unlocks
* Emissions rate
* Staking or reward dilution
* Treasury control
* Real utility versus forced token demand
The key question is simple: **who gets paid if this project succeeds, and who gets diluted while waiting?**
Bad tokenomics often looks like this:
* Low float, high FDV launch
* Large insider allocations with weak lockups
* Constant sell pressure from emissions
* Token required for narrative reasons, not product reasons
Good tokenomics does not guarantee upside, but bad tokenomics can cap upside for a very long time.
## What social and community data should you check? [#what-social-and-community-data-should-you-check]
You should check whether social activity reflects real interest or temporary hype. Social data helps you understand attention, but attention alone is not conviction.
Useful questions:
* Is engagement growing steadily or spiking unnaturally?
* Are people discussing product usage, or only price?
* Is the community broad, or driven by a few loud accounts?
* Is sentiment improving while on-chain usage also improves?
Social signals are most useful when combined with fundamentals.
For example:
* **Strong social growth + rising on-chain activity** can support a real breakout
* **Explosive social growth + weak usage** often signals speculation first, fundamentals later
* **High hype + concentrated holders** is where risk gets dangerous
That is why experienced investors do not stop at sentiment. They use it as context, not proof.
## What on-chain metrics should you check before buying a coin? [#what-on-chain-metrics-should-you-check-before-buying-a-coin]
You should check whether the network or token shows real usage, healthy participation, and behavior that matches the narrative. On-chain metrics are useful because they are harder to fake than marketing.
Depending on the asset, useful signals include:
* Active addresses
* Transaction count
* Transfer volume
* Holder growth
* Smart money or whale flows
* Treasury changes
* Protocol TVL, fees, or revenue
* Stablecoin inflows or outflows
The point is not to check every metric for every asset. The point is to check the metrics that match the business model.
Examples:
* For an L1, you care about active addresses, transactions, stablecoin activity, and ecosystem depth.
* For a DeFi protocol, you care about TVL, fees, revenue, user retention, and liquidity.
* For a meme coin, holder concentration and liquidity may matter more than almost anything else.
For DeFi-specific checks, sources like [DefiLlama](https://defillama.com) are useful because they surface TVL, fees, revenue, DEX volume, and token liquidity in one place.
This is also where manual research becomes heavy. One serious investment decision can send you across explorers, market sites, token unlock dashboards, audit pages, social feeds, and protocol analytics just to form a basic view.
## What is a practical crypto buying checklist? [#what-is-a-practical-crypto-buying-checklist]
A practical crypto buying checklist is a repeatable set of questions you answer **before** entering any position. If you cannot answer most of these, you probably are not ready to buy.
Use this list:
* What problem does the project solve?
* What is the current market cap and FDV?
* Is 24h volume healthy and consistent?
* Is liquidity deep enough for my position size?
* Who holds the supply?
* Are insiders locked or already liquid?
* Has the smart contract been audited?
* Are there upcoming unlocks or emissions that could pressure price?
* Is social attention matched by real on-chain activity?
* Are whales accumulating or distributing?
* Does the current market environment support risk-taking?
* What would make me wrong after I buy?
That last question matters most. Good investors do not just build a buy case. They build an invalidation case.
If you want a broader approach to market timing, combine project-level research with higher-level signals like [Bitcoin dominance](/blog/bitcoin-dominance-explained) and [market indicators](/docs/classic/market-indicators).
## How can you speed up crypto due diligence without skipping important checks? [#how-can-you-speed-up-crypto-due-diligence-without-skipping-important-checks]
You speed it up by using a consistent checklist and centralizing the inputs, not by skipping the hard parts. The goal is not to do less thinking. The goal is to waste less time collecting scattered data.
This is exactly why BlockMind exists. On Pro, BlockMind gives you a **personal AI investing agent** with its own workspace. Ask it to research a token before you buy, and it works through the same checklist using project fundamentals, market data, holder distribution, technical levels, sentiment, and risk. If you have connected a portfolio, the report can account for what you already hold.
The product turns that work into four clear steps: Explore scans the market for ideas, Analyze turns one into an analysis report, Verdicts asks a panel of experts for a call of approved, watch, wait, or reject, and Track keeps watching the assets you follow. Material changes can appear in your Morning Brief on the dashboard and in your email inbox. If you only want a one-off report, run a free [DeepDive analysis](https://blockmind.app/deepdive) first.
One boundary is deliberate: this is research, not financial advice. Your agent will never tell you what to buy or sell. BlockMind uses wallet and exchange connections only to read balances and positions, and the agent cannot trade, withdraw, or move funds even if you told it to.
That does **not** mean you should outsource judgment. It means you should stop wasting energy jumping between ten tabs just to assemble the first draft of your view.
For a bigger-picture look at that tradeoff, see [AI vs Manual Crypto Research](/blog/ai-vs-manual-crypto-research).
## Frequently Asked Questions [#frequently-asked-questions]
### What to check before buying crypto? [#what-to-check-before-buying-crypto]
Check market cap, FDV, trading volume, liquidity, holder distribution, team credibility, audits, tokenomics, social activity, and on-chain usage before buying crypto. Looking at price alone is not enough.
### What is the best crypto buying checklist? [#what-is-the-best-crypto-buying-checklist]
The best crypto buying checklist is one you can repeat on every asset. It should cover valuation, liquidity, ownership concentration, team quality, security, tokenomics, sentiment, and on-chain activity.
### How do I know if a cryptocurrency is safe to buy? [#how-do-i-know-if-a-cryptocurrency-is-safe-to-buy]
You never know with certainty that a cryptocurrency is safe to buy. What you can do is reduce avoidable risk by checking audits, holder concentration, liquidity, token unlocks, team credibility, and on-chain behavior before entering a position.
### What on-chain metrics matter before buying a coin? [#what-on-chain-metrics-matter-before-buying-a-coin]
The most useful on-chain metrics depend on the asset, but common ones include active addresses, transaction count, holder growth, whale flows, TVL, fees, revenue, and liquidity. The right metric is the one that matches how the project is supposed to create value.
### How much research should you do before buying a cryptocurrency? [#how-much-research-should-you-do-before-buying-a-cryptocurrency]
You should do enough research to explain in plain language what the project does, how the token works, what the main risks are, who owns the supply, and what would make your thesis wrong. If you cannot do that, you probably should not buy yet.
## The Bottom Line [#the-bottom-line]
Buying crypto responsibly is less about finding a secret indicator and more about checking the obvious things that most people skip. Market cap, volume, liquidity, holder distribution, audits, tokenomics, social activity, and on-chain metrics will not eliminate risk, but they will filter out a lot of bad decisions.
That is also the uncomfortable truth: good crypto research is real work. If you do it properly, it takes time.
## Sources [#sources]
1. [Chainalysis: 2026 Crypto Scam Trends](https://www.chainalysis.com/blog/crypto-scams-2026/), January 2026.
2. [CoinGecko: Global Crypto Market Charts](https://www.coingecko.com/en/global-charts), accessed July 2026.
3. [DefiLlama: DeFi Data and Methodology](https://defillama.com), accessed July 2026.
4. [SEC Investor.gov: Crypto Asset Scam Warning Signs](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/crypto-scams), May 2024.