Inspect a decision

An Actera decision preserves the runtime authorization result for one proposed AI agent action and separates that result from later approval and execution state. Open Decisions, choose a row, and verify the evidence in this order:

1. Agent and action identify who acted and what it intended to do.
2. Outcome and state distinguish the policy effect from the current lifecycle state.
3. Evaluation context preserves the exact facts Actera evaluated.
4. Policy evidence identifies every applicable policy and immutable version.
5. Rule trace shows conditions, contributing effects, and precedence.
6. Approval evidence identifies the approving or rejecting actor and timestamp.
7. Activity and execution evidence records what the trusted integration reported afterwards.

Approval alone does not mean the action executed. Look for a completed execution report or subsequent activity. Treat an unknown outcome as a reconciliation task, not permission to blindly retry.

For automation, use Retrieve a decision. Review Decisions and approvals for the full lifecycle and How Actera works for effect precedence.