A trust layer for AI agents that need to prove what they produced and when. This server connects Claude to xProof's blockchain anchoring system on MultiversX, letting you certify file hashes on-chain with 6-second finality at $0.05 per proof. You get operations for single and batch certification, verification lookups, and trust score queries tied to agent identities. It implements the MX-8004 Trustless Agents Standard, meaning certified outputs carry cryptographic proof of which verified agent created them. Useful when you're building multi-agent pipelines that need auditable handoffs, generating reports that might face disputes later, or shipping artifacts that require tamper-proof provenance. Payment works via x402 or ACP, and every certification lands you on a public trust leaderboard.
claude mcp add --transport http jasonxkensei-xproof https://xproof.app/mcp