Skip to main content

​
For developers

Two-line integration — works with CrewAI, LangGraph, LangChain, AutoGen, or any custom agent framework. No token management — the SDK caches tokens, renews them before expiry, and detects revocations automatically. Just call get_token(). Test before you ship — the Policy Simulator lets you test access rules against mock requests before they go live. AI policy writing — describe what you want in plain English and Anti AI generates the Rego policy for you.

​
For security teams

Every agent has its own identity — no more shared API keys. Revoke one agent without touching others. Instant revocation — revoke any agent or token from the dashboard. The SDK detects it within 15 seconds. Blast Radius view — see exactly what tools and services each agent can reach if compromised. Geofencing — restrict agent token usage to specific geographic regions. Ghost agent detection — discover unregistered AI workloads making requests through your infrastructure.

​
For compliance teams

Tamper-proof audit trail — every action logged with a cryptographic hash chain. Any modification is immediately detectable. One-click SOC 2 reports — generate a cryptographically signed report covering any time range. Human accountability mapping — every agent action is traceable back to the human who owns that agent. Export anytime — CSV, JSON, or NDJSON. All exports are signed. 365-day retention — logs kept for a full year by default.