Check AI Agent Trust Before Invocation

Check an AI Agent Before You Invoke It

Paste the exact endpoint URL your agent is about to call. AgentTrust returns a trust decision from the monitoring, reliability and ownership evidence it already holds — without contacting the target.

How it works

01Looks the URL up among the public agents AgentTrust already observes — an exact match on the invocation URL.

02Reads that agent's existing monitoring history and endpoint ownership verification — not a live probe.

03Returns a machine-readable trustDecision — recommended, confidence and reasons — for you to act on.

AgentTrust never contacts the target endpoint

A trust check reads only what AgentTrust has already observed — past health checks and verification state — so it's safe to run before you've decided the endpoint is worth talking to.

For AI agents and MCP clients

The same check is the anonymous, read-only MCP tool check_agent_trust at https://getagenttrust.com/api/mcp (Streamable HTTP).

No AgentTrust account or API key

Exact match on the URL you supply

Unknown URL → { matched: false }, never an error

Rate-limited per caller IP

check_agent_trust({ endpointUrl: "https://agenttrust-umber.vercel.app/api/test-agent" })

Example result — fictional values

{
  "matched": true,
  "slug": "example-agent",
  "name": "Example Agent",
  "status": "healthy",
  "verified": true,
  "reliabilityScore": 96,
  "reliabilityScoreStatus": "fresh",
  "trustDecision": {
    "recommended": true,
    "confidence": "high",
    "reasons": []
  }
}

What this is, and isn't

The trustDecision is derived from an agent's monitoring history, reliability score and endpoint ownership verification — signals AgentTrust has itself observed over time. It is not a community reputation or rating system, offers no security guarantee, and does not prove an agent is safe to use: it reports what AgentTrust has observed so you can make your own decision.