TrustSniffer MCP server
Ask your AI assistant (ChatGPT, Claude, Cursor, VS Code or any other app that supports MCP) whether a website or a crypto wallet is safe, and get TrustSniffer's real analysis back in the chat. It runs on your TrustSniffer account, with the same checks, prices and daily energy as the website.
https://trustsniffer.com/mcp
Step 1What it does
MCP (Model Context Protocol) is the open standard AI assistants use to call outside tools. Once you connect TrustSniffer, your assistant can:
- Read a website's trust report: the score from 0 to 100 (higher is safer), the verdict, what to do and the key findings, with a link to the full report.
- Analyse a website that has no report yet, or a fresh one: page content, behaviour in a sandboxed browser, domain age, web archive and reputation services.
- Check a crypto wallet on Ethereum or TRON for sanctions, USDT/USDC freezes, known scam and theft wallets, and exposure to them.
- Show your energy: what is left today and what each action costs.
Every answer comes from TrustSniffer's own analysis engine. The assistant does not invent scores: it quotes ours and links the report.
Step 2Connect it
The server works with any AI app that can connect to a remote MCP server. Every app asks for the same thing: the server URL above. Then a TrustSniffer page opens where you sign in and press Approve. Menu names change from one app version to the next; if yours looks different, look for "connectors", "MCP servers" or "tools" in its settings.
ChatGPT
- Open SettingsApps & ConnectorsAdvanced settings and turn on Developer mode.
- Create a connector: name it
TrustSniffer, paste the server URL and choose OAuth sign-in. - Approve the connection on the TrustSniffer page that opens. Custom connectors need a ChatGPT plan that includes developer mode.
Claude (web, desktop and mobile)
- Open SettingsConnectorsAdd custom connector
- Name it
TrustSniffer, paste the server URL and click Connect. - Approve the connection on the TrustSniffer page that opens.
Cursor
Add this to ~/.cursor/mcp.json (or .cursor/mcp.json in a project), then press Connect next to TrustSniffer in Cursor's MCP settings:
{
"mcpServers": {
"trustsniffer": { "url": "https://trustsniffer.com/mcp" }
}
}
VS Code (GitHub Copilot)
Run MCP: Add Server from the Command Palette, choose HTTP and paste the server URL. Or add it to .vscode/mcp.json:
{
"servers": {
"trustsniffer": { "type": "http", "url": "https://trustsniffer.com/mcp" }
}
}
Command-line assistants
# Claude Code (then run /mcp and choose Authenticate)
claude mcp add --transport http trustsniffer https://trustsniffer.com/mcp
# Gemini CLI (then run /mcp auth trustsniffer)
gemini mcp add --transport http trustsniffer https://trustsniffer.com/mcp
Any other app
If your app supports remote MCP servers over Streamable HTTP with OAuth sign-in, it works the same way: give it the server URL and approve the connection when TrustSniffer asks. Apps that only run local MCP servers cannot connect.
Step 3Who can connect
A free TrustSniffer account is enough, with two conditions that keep throwaway accounts out:
- your email address is verified, and
- the account is at least 3 days old.
Until then the website works as usual; the approval page tells you the exact date connections open for your account. Create an account
Step 4The tools
| Tool | What it does | Cost |
|---|---|---|
get_website_report | The published trust report for a website. | Free, up to 50 a day per account |
analyze_website | Runs a new website analysis and publishes the report. Your assistant asks you before running it. | Website analysis price in energy |
get_analysis_status | Progress of an analysis that is still running. | Free |
check_wallet | AML screen of an Ethereum or TRON wallet. Your assistant asks you before running it. | Wallet check price in energy |
get_energy_balance | Your remaining energy today and the price of each action. | Free |
The reading tools are marked read-only, so an assistant may run them without asking. The two tools that spend energy are marked as actions, so an assistant asks for your permission first.
Step 5Energy and limits
One account, one daily energy
The connection spends your account's own daily energy, at the same prices as the website, the extension and the API. There is no separate allowance and no way to spend more than you have. Energy resets every day at 00:00 UTC.
- An analysis that ends as Needs Review, or a recent report that already exists, is not charged.
- A wallet check that cannot be completed is not charged and is never reported as clean.
- Reading reports is free and limited to 50 a day per account through MCP. Reports stay free to read on the website without limit.
- The free daily energy is one per person. A second account made in the same browser does not get its own.
Your balance and the price of each action are shown in your dashboard.
Step 6Security and privacy
- You never give your password to the AI app. You sign in on trustsniffer.com and approve the connection there (OAuth 2.1 with PKCE). The approval page shows which app is asking and where it will send you back.
- The app receives a token that expires after an hour and is renewed automatically. If a renewal token is ever used twice, the connection is closed at once.
- You can see and disconnect every connection in DashboardSettingsAPI & DevelopersAI connections. Deleting your account removes them all.
- TrustSniffer receives only what a tool is asked: a website address or a wallet address. It never sees your conversation, your memory or your files.
- We store the connection's app name, when it was made and last used, and how many reports it read today. Analyses you start are published like any other TrustSniffer report. Details: Privacy policy.
Step 7Things to ask
Step 8Questions
- Is the MCP server free?
- Yes. Connecting is free and reading reports is free. Analyses and wallet checks use your daily energy, which every account receives free each day.
- Which AI apps work?
- Any AI app that supports remote MCP servers with OAuth sign-in, including ChatGPT, Claude, Cursor, VS Code with GitHub Copilot, Claude Code and Gemini CLI. TrustSniffer is not tied to one AI company: the same server and the same account work in all of them.
- Why does my account need to be 3 days old?
- So that nobody can create throwaway accounts and plug them into scripts or assistants to collect free energy. The website itself works from the first minute.
- What do the scores mean?
- The website trust score runs from 0 to 100, and higher is safer: 75 and above is High Trust, 50 to 74 Moderate, 25 to 49 Low and under 25 Critical. The wallet risk score runs the other way: 100 is the riskiest. A score rates the evidence TrustSniffer found; it is not a guarantee.
- How do I disconnect?
- In Dashboard → Settings → API & Developers → AI connections, or by removing the connector in your AI app. Either one ends the connection.
- Who do I contact?
- [email protected]
Prefer code? The developer API offers the same checks with a personal API key.