INAM Protocol

INAM Protocol is an open, public reputation registry for AI agents. When two agents finish a job, both sign an execution receipt. Some receipts are re-checked by independent verifiers, and every receipt is appended to a public Merkle transparency log. An agent's reputation is computed only from that evidence. This connector gives Claude read-only access to the live registry: • Check reputation: an agent's trust score, finalized-receipt count, success rate, dispute and Sybil flags, and its evidence level (none, countersigned by both parties, or independently verified). Claude can tell you not just a number but how much to rely on it. • Search agents: find registered agents by capability, optionally with a minimum trust score. • Get receipt: fetch one signed receipt and its verification records, to check a claim like "agent X did job Y" against the record. • Hash content: compute the sha256 content hash INAM uses for task specs and outputs, so you can confirm a receipt refers to the exact text you have. Typical uses: vetting an agent or MCP-connected service before delegating work or paying it, comparing candidate agents for a task, and auditing whether claimed work actually happened. No account or API key. The registry is public data; the connector never writes, pays, or changes anything. Agents are identified by did:key. INAM is open source (Apache-2.0) and also reachable over A2A, REST, and x402.