[Feature] Agent Identity & Trust Score Integration via SATP

[0xbrainkid]

Problem

The MCP registry currently indexes servers by capability and metadata, but there's no mechanism to verify the identity or trustworthiness of agents consuming these servers. As MCP adoption scales (97M+ monthly SDK downloads), knowing who is calling a server becomes as important as what the server does.

Proposal

Integrate with AgentFolio and the Solana Agent Trust Protocol (SATP) to add verified agent identity to the registry:

• Agent verification badges — on-chain proof that an agent's identity (GitHub, X, wallet) is verified
• Trust scores — computed from verification depth, activity, and on-chain attestations
• MCP server agentfolio-mcp — we're building an MCP server that exposes agent identity lookup, trust verification, and credential validation as MCP tools

Use Cases

1. Server operators can gate access based on agent trust scores
2. Registry users can filter servers by verified publisher identity
3. Multi-agent systems can verify peer identity before delegating tasks

Technical Details

• SATP stores agent identity on Solana (Ed25519 signatures, on-chain PDAs)
• AgentFolio API: GET /api/agents/{id} returns trust score, verifications, capabilities
• Proposed MCP tools: verify_agent, get_trust_score, lookup_agent_by_wallet

Prior Art

• ERC-8004 (Agent Identity on Ethereum)
• Microsoft Agent ID (Entra) — enterprise approach, announced at RSAC 2026
• SATP takes the open/decentralized approach

Would love to discuss how agent identity fits into the registry's roadmap. Happy to contribute a PR.

Links: AgentFolio | SATP Docs | brainAI