Lelu - Authorization Engine for AI AgentsLelu.

Introducing|Authorization Protocol<br>The authorization layer for autonomous AI agents<br>Get StartedTry Sandbox

Community/Changelog/Legal/Careers

READMEDOCSSDKENTERPRISECHANGELOG

README<br>Authorization that lives inside your agent . Confidence-aware, policy-driven, and built to scale — from weekend AI projects to production agent swarms.<br>CLITypeScriptcurl<br>Copy<br>npm install lelu-agent-auth<br>npx lelu init<br>Trusted by teams building with<br>Vercel AI SDKLangChainOpenAIAnthropicClaudeMistralLlamaIndexCrewAIAutoGPTVercelFly.ioRailway

Features<br>01<br>Confidence-aware gating.<br>Every agent action carries a confidence score. Low confidence routes to human review automatically — no code change needed.

02<br>Human-in-the-loop review.<br>Agents pause, queue their action, and resume once a human approves or denies. Full audit trail included.

03<br>Policy-driven authorization.<br>Write Rego policies to express complex rules. Allow, deny, or require review — per actor, action, or resource.

QuickstartAPI ReferenceSDK DocsGitHub