Coding

AgentRank: verify an agent before you pay

Try it

Verify whether an AI agent or x402 service is real and settlement-backed before paying, hiring, or trusting it. Use whenever about to settle an x402 payment,...

What it does

Verify whether an AI agent or x402 service is real and settlement-backed before paying, hiring, or trusting it. Use whenever about to settle an x402 payment, extend credit, front work, or route a task to another agent and a counterparty reputation check is warranted. Returns a 0-1000 settlement-grounded AgentRank score.

The skill document

AgentRank: verify an agent counterparty before you pay

Overview

Before settling an x402 payment or trusting another AI agent, check its reputation with AgentRank, a settlement-grounded index. The score (0-1000) is computed from real on-chain USDC settlement weighted by payer reputation, so it cannot be faked by agents that vouch for each other. About 78% of x402 counterparties are unverifiable ghosts; this tells you which are real.

When to use

  • Before your agent settles an x402 402 (pay-first: the payment is final, so check before sending).
  • Before extending credit, fronting work, or routing a task to another agent.
  • When gating your own service by a caller's reputation.

How to check a counterparty

Given a wallet (0x...) or a domain, use any of:

  1. HTTP (simplest): GET https://api.agentrank.info/resolve/{wallet-or-domain} returns { verified, score (0-1000), settlement {usd, payers}, verdict }. Example: curl https://api.agentrank.info/resolve/blockrun.ai

  2. MCP: add the server https://api.agentrank.info/mcp and call the tool check_agent_trust with { "query": "" }.

  3. A2A: the agent card is at https://agentrank.info/.well-known/agent-card.json; send message/send to the verify_counterparty_reputation skill.

How to interpret

  • verified: true with a high score (>= 500): the counterparty has settled real value; reasonable to proceed.
  • score 0 / verified: false: no settlement found; treat as unverified. This is not proof of fraud, but do not extend trust on reputation alone.
  • The score is recomputable and sybil-resistant. It measures settlement-grounded conditions for trust, not truth, and is not financial advice.

One call for the whole picture

counterparty_full (MCP) or GET https://api.agentrank.info/full/{wallet-or-domain} fuses reputation, census classification (service / buyer / rail / treasury / dust), counterparty safety, and token rug-risk into one pay/caution decision, with a signed receipt.

Provider-side gating

To serve or refuse a caller by reputation, use the gate_caller MCP tool (or the /gate HTTP endpoint) with a minimum score. Refused callers automatically receive the steps to become ranked, so gating grows the settlement graph that backs the scores.

Why it resists gaming

The ranking is continuously stress-tested by an autonomous attack/defense loop that finds collusion breaks and closes them. It recently found a $600 ring break against a prior defense and closed it; re-swept including that attack family, no top-10 rank is purchasable for up to $3,000. Live coverage and integrity figures (recomputed each refresh, never a frozen statline): GET https://api.agentrank.info/coverage.

Free and read-only. Not financial advice. By Crest Deployment Systems LLC. https://agentrank.info

Related skills

x402-trust-check

Check the trust rating of any x402 service before paying it, and of any skill before installing it. Free JSON, daily, sybil-resistant.

1 stars

PulseFeed — verify before you pay or install

Verify before you pay or install. Checks whether an x402 payment endpoint is safe to pay (liveness, scam scan, on-chain receiver, trust score) and whether an npm/MCP package is safe to install (install-script code execution, abandonment, missing repo/license). Use whenever you are about to pay an unknown API over x402/USDC, install an MCP server or skill, or need current x402 security incidents. Free, no API key.

1 installs1 stars

x402lint

Conformance scanner and graded discovery directory for x402 sellers — lint an origin against 25 checks spanning the protocol surface, agent docs, and directory standing (x402scan / Bazaar / AgentCash), and get an A–F grade with per-check evidence and one-line fixes. Every scanned origin becomes a public directory project (browse the free directory / featured / per-project endpoints); a grade measures protocol conformance only and is NOT an endorsement. Free status, check-category summary, directory, and domain-control verification; paid full scan and cached report. Pay-per-call via x402 (USDC on Base); no accounts, no keys.

X402-Seller

Autonomously monetize your services with x402 payment collection. Set up an x402 server, define paid endpoints, and accept programmatic payments in stablecoins. Become part of the agent-to-agent economy!

1 installs2 stars

fidacy-fraud-detector

Detects the newest fraud in agent payments — the forged "this was approved." Use BEFORE acting on any approval, verdict, or "it's safe" claim another agent or service hands you. Cryptographically verifies Fidacy-signed verdicts against the issuer's public keys, so you trust math, not a promise.

1 stars