Foo Guard overview
Learn what Foo Guard analyzes and how it helps secure AI agents before production.
Foo Guard analyzes AI agent identity, authorization, tools, credentials, autonomy, and auditability.
The deterministic rules engine is authoritative for severity, scoring, and pass/fail findings. Optional AI features may explain findings and suggest remediation, but they do not decide whether a vulnerability exists.
Core workflow
1. Provide an agent configuration 2. Foo Guard normalizes and analyzes it 3. Deterministic rules produce findings and a score 4. You review remediation guidance and fix misconfigurations
Product areas
- Personal workspace for individual assessments and saved analyses
- Team workspaces for organization governance, GitHub scanning, policy, and audit history
- CLI and API for CI/CD and programmatic analysis