Browse AI Agents & MCP SECTION 8: Safety, Guardrails & Human-in-the-Loop

The Agent Audit Trail Designer

2 fill-in slots · from The AI Prompt Handbook for AI Agents & MCP

Your details

0/2 filled

The Agent Audit Trail Designer

Design an audit trail for this agent.

Agent: [DESCRIBE]
Regulatory context: [DESCRIBE, OR "none stated"]
Who needs to review it: [DESCRIBE]

Produce:
1. What must be recorded for every action with a consequence
2. The record structure: who, what, when, why, on whose behalf,
   with what authority
3. How the agent's reasoning is captured, to the extent it is
   available
4. How to link an action back to the human request that caused it
5. What is captured about approvals: who approved, what they saw,
   when
6. Immutability: what prevents the record being altered
7. Retention and disposal
8. What must be redacted, and how
9. How a reviewer reconstructs "why did this happen" from the trail

Then:
10. If someone asked "prove this agent did not do X", could you?
11. If a customer asked "why did your system do this to my
    account", could you answer completely?

Mark all regulatory specifics as [VERIFY AGAINST APPLICABLE
REGULATION AND LEGAL ADVICE].

Highlighted [slots] are still empty. Downloads are plain .txt — paste into any AI assistant.