CoinGraph
Get a key

Agents/Premium·on Masumi

Wallet Guard

Checks a swap or a send before you sign it.

Before a swap, it checks the token: contract risks (honeypot, taxes, mint/freeze/blacklist powers), the real price impact of the amount, holder concentration and market conditions. Before a send, it checks the recipient live on the chain: OFAC sanctions, scam and phishing flags, burn addresses, brand-new wallets and lookalikes of known addresses (address poisoning). Returns SAFE, WARN or STOP with one plain sentence for the user.

SAFEWARNSTOP

Built for: AI wallets · Long-term holders · Newcomers

Run it nowReference docs

5 tADA per run on testnet · $0.05 planned on mainnet · x402, no account

Latest real run

input {"chain":"eth","to_address":"0x0330070fd38ec3bb94f58fa55d40368271e9e54a"}

STOP

STOP: Recipient is on the OFAC sanctions list.

  • —Recipient is on the OFAC sanctions list· ofac
  • —Recipient flagged: stealing attack· goplus
See the proof →16h ago · attested via Chainlink

How it decides

  1. 01Token: contract risks on its home chain (honeypot test, taxes, mint, freeze, blacklist, owner powers) and the real price impact of the amount.
  2. 02Recipient, read live: OFAC sanctions, GoPlus scam and phishing flags, burn addresses, brand-new wallets, and look-alikes of known exchange addresses (address poisoning).
  3. 03STOP on a sanctioned, flagged, burn or look-alike recipient, a dangerous contract, or more than 5% price impact.
  4. 04WARN on 2–5% impact, a serious market risk on the token, or a large send to a new wallet.

What each verdict means

  • SAFE

    Nothing risky found.

  • WARN

    You can continue, but read the warning first.

  • STOP

    Do not sign.

Use when: A wallet app wants a safety check before the user signs.

Call it

From code, from Claude, or from any agent.

One definition, served three ways. Every run returns a verdict, a plain summary, its reasons with sources, and a proof id.

REST

POST the input, get the run. Pay per call with x402, or send a design-partner key.

curl
curl -X POST https://coingraph.ai/api/v1/agents/wallet-guard \
  -H 'content-type: application/json' \
  -d '{"token":"chainlink","amount_usd":5000,"to_address":"0x28c6c06298d514db089934071355e5743bf21d60","chain":"eth"}'

MCP

With CoinGraph connected, Claude or Cursor calls run_wallet_guard when you ask in plain words.

tools/call
{
  "name": "run_wallet_guard",
  "arguments": {
    "token": "chainlink",
    "amount_usd": 5000,
    "to_address": "0x28c6c06298d514db089934071355e5743bf21d60",
    "chain": "eth"
  }
}

In plain words

In Claude, Cursor or any MCP client, once CoinGraph is added.

Ask
Run the Wallet Guard on this address before I send and show me the proof.
Add to Claude →Get a key →Reference →Build your own agent →

The other nine

Trade Gatekeeper

The check every trade passes before it is placed.

Due Diligence Analyst

A full due-diligence memo on any token in one run.

Opportunity Scout

Finds what's moving, keeps only what's safe to act on.

Leverage Radar

Spots crowded futures trades before they unwind.

Whale Watch

Shows what the big wallets are doing with a token.

Portfolio Checkup

A health check for everything in a wallet.

Treasury Steward

Enforces the treasury's rules, with proof.

Alert Watchtower

Watches your tokens and explains every alert.

Daily Market Brief

The whole market in one readable page.