AgentOS - AI Agent Governance Control Plane

Gallery

About

AgentOS is a self-hosted control plane for autonomous AI agents: every action an agent wants to take is submitted for a policy decision (ALLOW / APPROVAL_REQUIRED / BLOCK) before it can reach a real tool.What's included: full source code (Node.js/TypeScript backend, React + Vite console); a pure policy engine with mandate-based decisioning; a gateway that is the sole path to tool execution, with full audit logging; role-based access control (viewer / operator / approver / admin); webhook SSRF protections and hardened security fixes; mutation testing, chaos/fault-injection testing and a full test suite; complete documentation (architecture, setup, deployment).Tech stack: Express, Drizzle ORM, SQLite, React, TanStack Query, Tailwind, Zod, Vitest.Ideal for: AI security startups, teams deploying AI agents in production, and companies that need governance and compliance over agent actions.Non-exclusive license: full working codebase to deploy, white-label and build on. Pre-revenue (code + documentation only, no users). Optional maintenance and support available. Open to offers.