Govern API access for AI agents, detecting breaking changes and blocking unsafe calls.
Run risk-aware API preflight checks for safer agent calls and merges.
