Bissap: the bouncer for your app

Your users already log in the normal way. Bissap lets them also bring their AI agent. The agent gets its own key, its own permissions, and its own limits. If it behaves, it gets in. If it doesn't, it gets kicked out.

You decide what agents are allowed to do. Your users decide if they want to use one.

One line of code adds a full agent API to your server: key minting, scoped permissions, rate limiting, intent routing, and audit logging. Works with any Node.js or Python backend.

What Bissap gives you

🔑 Keys: per-user, scoped, revocable. Your users mint a key for their agent. You control what it can access.

🎯 Intents: named actions with schemas and scope enforcement. Define what agents can do, and the SDK enforces it automatically.

📨 Events → Inbox: your app sends events, agents poll and ack. Like push notifications, but for AI agents.

🛡️ Cloud dashboard: revoke keys, track usage, set auto-revoke policies, view audit logs. The SDK handles enforcement locally; Cloud handles coordination globally.

Get started

➡️ Quickstart (~15 min): go from zero to a working agent API

📚 Mental model: understand how Bissap fits into your app

🔧 CLI reference: manage projects from the terminal