@getaid/mcp-server

MCP (Model Context Protocol) server for querying the getaid.dev universal AI agent identity registry.

Enables any MCP-compatible AI agent or client (Claude, etc.) to look up agent identities, verify trust levels, and perform cryptographic identity challenges against the getaid.dev registry.

Installation

Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "getaid": {
      "command": "npx",
      "args": ["-y", "@getaid/mcp-server"],
      "env": {
        "GETAID_API_KEY": "your-api-key-here"
      }
    }
  }
}

Claude Code

claude mcp add getaid -- npx -y @getaid/mcp-server

Manual / Other MCP Clients

npx @getaid/mcp-server

Configuration

| Environment Variable | Description | Default | | -------------------- | ------------------------------ | ------------------------- | | GETAID_API_URL | Base URL of the getaid.dev API | https://getaid.dev | | GETAID_API_KEY | API key for higher rate limits | None (public rate limits) |

Tools

lookup_agent

Look up a full agent profile by AID or handle.

| Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------------------------------ | | aid | string | No* | Agent ID (e.g. AID-000012347) | | handle | string | No* | Agent handle (e.g. @johndoe/trading-bot) |

* At least one of aid or handle must be provided.

search_agents

Search the agent directory with filters.

| Parameter | Type | Required | Description | | ------------------ | ------ | -------- | ------------------------------------------------------------ | | query | string | Yes | Search query | | status | string | No | Filter by status (comma-separated) | | verificationTier | string | No | Filter by tier (comma-separated) | | capabilities | string | No | Filter by capabilities (comma-separated) | | tags | string | No | Filter by tags (comma-separated) | | protocols | string | No | Filter by protocols (MCP, A2A, REST, GraphQL, gRPC) | | sortBy | string | No | relevance, newest, verification_tier, last_heartbeat | | limit | number | No | Results per page (1–100) | | cursor | string | No | Pagination cursor |

verify_agent

Quick trust check — returns AID validity, verification tier, and status.

| Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------ | | aid | string | Yes | Agent ID to verify |

get_agent_status

Get heartbeat/liveness status of an agent.

| Parameter | Type | Required | Description | | --------- | ------ | -------- | ----------- | | aid | string | Yes | Agent ID |

get_public_key

Get an agent's Ed25519 public key for signature verification.

| Parameter | Type | Required | Description | | --------- | ------ | -------- | ----------- | | aid | string | Yes | Agent ID |

create_challenge

Issue a cryptographic identity challenge. Returns a nonce the agent must sign.

| Parameter | Type | Required | Description | | --------- | ------ | -------- | --------------------- | | aid | string | Yes | Agent ID to challenge |

verify_challenge

Verify a signed challenge response to confirm agent identity.

| Parameter | Type | Required | Description | | ----------- | ------ | -------- | ----------------------------- | | aid | string | Yes | Agent ID | | nonce | string | Yes | Nonce from create_challenge | | signature | string | Yes | Hex-encoded Ed25519 signature |

Cryptographic Verification Flow

To verify an agent's identity:

1. Call create_challenge with the agent's AID → receive a nonce
2. The agent signs the nonce with its Ed25519 private key
3. Call verify_challenge with the AID, nonce, and signature → receive verification result

Development

# Run tests
pnpm test

# Type check
pnpm typecheck

# Build
pnpm build

License

MIT