AgentMesh API

Public API for AgentMesh trust verification and governance services.

Overview

This API enables AI agents to:

• Register with the AgentMesh network
• Perform cryptographic trust handshakes
• Verify other agents' trust scores
• Query audit logs

Endpoints

Method	Endpoint	Description
GET	/api	Health check and stats
POST	/api/register	Register a new agent
GET	/api/verify/:did	Verify an agent's status
POST	/api/handshake	Perform trust handshake
GET	/api/score/:did	Get detailed trust score
GET	/api/audit/:did	Query audit log

Quick Start

Register an Agent

curl -X POST https://agentmesh-api.vercel.app/api/register \
  -H "Content-Type: application/json" \
  -d '{
    "name": "MyAgent",
    "sponsor_email": "owner@example.com",
    "capabilities": ["read", "write"]
  }'

Verify an Agent

curl https://agentmesh-api.vercel.app/api/verify/did:mesh:abc123

Trust Handshake

curl -X POST https://agentmesh-api.vercel.app/api/handshake \
  -H "Content-Type: application/json" \
  -d '{
    "agent_did": "did:mesh:abc123",
    "challenge": "random_challenge",
    "capabilities_requested": ["read", "execute"]
  }'

Development

# Install dependencies
npm install

# Run locally
vercel dev

# Deploy
vercel deploy

Environment Variables

For production deployment, configure:

• KV_REST_API_URL - Vercel KV endpoint
• KV_REST_API_TOKEN - Vercel KV token

Without these, the API uses in-memory storage (data resets on redeploy).

Architecture

agentmesh-api/
├── api/                  # Vercel serverless functions
│   ├── index.ts          # Health check
│   ├── register.ts       # Agent registration
│   ├── verify.ts         # Agent verification
│   ├── handshake.ts      # Trust handshakes
│   ├── score.ts          # Trust scores
│   └── audit.ts          # Audit logs
├── lib/                  # Shared libraries
│   ├── types.ts          # TypeScript types
│   ├── storage.ts        # Vercel KV / memory storage
│   ├── crypto.ts         # Cryptographic operations
│   └── trust.ts          # Trust score calculations
└── public/               # Static files
    ├── skill.md          # Moltbook skill definition
    └── heartbeat.md      # Status page

Related Projects

• agent-mesh - Core AgentMesh Python library
• agent-os - Agent Operating System kernel
• agentic-architecture - Architecture documentation

License

MIT