› Documentation
OpenLeash documentation is maintained alongside the source code. Start here to find the right reference.
README & Quickstart
Project overview, installation, SDK usage, CLI commands, and architecture.
Protocol
Authorization request/response protocol, endpoints, and data formats.
Policy Language
How to write authorization policies: rules, conditions, obligations, and trust levels.
OpenClaw Integration
Run OpenLeash as a sidecar for OpenClaw. Configuration and hook points.
Agent Integration Guide
Step-by-step guide for AI agents: registration, authorization, and proof verification.
OpenAPI Reference
Interactive API documentation with Scalar UI. Full OpenAPI 3.0 spec for all endpoints.
Concepts
What is AI Agent Authorization?
Learn why AI agents need authorization, how policy-based governance works, and how cryptographic proofs enable trust.
MCP Authorization & Tool Governance
Understand the governance gap in MCP, how OpenLeash acts as an authorization sidecar, and how to set up policy-based tool control.
AI Agent Guardrails
How to set boundaries for autonomous agents — spending limits, action controls, approval workflows, and cryptographic accountability.
PASETO Proof Tokens
How PASETO v4.public tokens provide cryptographic proof of authorization. Why PASETO is more secure than JWT for agent proof.
Human-in-the-Loop AI
Approval workflows, step-up authentication, and configurable escalation policies for high-stakes AI agent actions.
OpenLeash vs API Keys
Why static API keys aren't enough for autonomous agents and when per-action authorization matters.
OpenLeash vs OAuth
How OpenLeash complements OAuth by adding per-action authorization on top of token-based delegation.
SDKs
TypeScript SDK
npm install @openleash/sdk-ts — authorization, signing, and proof verification for Node.js agents.
Python SDK
pip install openleash-sdk — authorization, signing, and proof verification for Python agents.
Go SDK
go get github.com/openleash/openleash/packages/sdk-go — for Go agents and microservices.
CLI commands
npx openleash start Start the sidecar (with GUI) npx openleash init Non-interactive setup npx openleash wizard Interactive setup npx openleash policy <cmd> list | show | upsert | validate | delete | unbind npx openleash owner <cmd> list | show | add-contact | add-gov-id | add-company-id | validate npx openleash keys <cmd> list | rotate npx openleash playground Test policies locally API endpoints
Public (no auth)
GET /v1/health Server health check GET /v1/public-keys Fetch signing keys for offline verification POST /v1/verify-proof Verify a proof token online GET /v1/reference OpenAPI reference (Scalar UI) Agent (Ed25519 signed)
POST /v1/authorize Request authorization + proof token POST /v1/agents/register Register agent with challenge proof GET /v1/agent/self Agent's own registration details POST /v1/agent/approval-requests Create approval request POST /v1/agent/policy-drafts Propose a new policy for owner review GET /v1/agent/policy-drafts List and poll policy draft status Owner (PASETO session)
POST /v1/owner/login Login with passphrase GET /v1/owner/agents List agents under owner GET /v1/owner/policies List owner's policies GET /v1/owner/approval-requests Pending + resolved approval requests POST .../approve Approve a request (issues approval token) GET /v1/owner/audit Owner-scoped audit log (filter by user/org/agent) GET /v1/owner/organizations/:orgId Organization details and members POST .../members/:userId Invite members, manage roles Admin (PASETO session with admin role)
GET /v1/admin/owners List all owners GET /v1/admin/organizations List all organizations GET /v1/admin/agents List all agents GET /v1/admin/policies List all policies GET /v1/admin/audit Full audit log GET /v1/admin/config Server configuration (read-only)