A record of how an agent behaved and what was done when its behavior moved outside expectations.
For a regulated business, catching a problem is only half the job. You also have to show what happened: what changed, when, which scenarios were affected, and what action followed. That record is what turns oversight into evidence, and it is often the most important thing the whole process produces.
Related: Agent Accountability, AI Agent Governance, Model Risk Management
See also: AI Agent Drift in Regulated Environments