Skip to main content

AI Governance

The AI control-plane: classify an action into its risk level, see the Policy Engine decision and whether it routes to human approval, manage the registry of callable tools, and review the append-only agent-action ledger. Classification is a read-only dry-run — no action is executed and no approval is created here.

Classify an action

Enter an action key to see its level, the Policy Engine decision, and whether it routes to approval. Unknown keys fail closed to blocked.

Action classification matrix

Tool registry

Recent agent actions