Audit & compliance

Audit trail for every AI agent action

When an AI agent deletes a production branch at 3 AM, the first question is always “what happened?” TameFlare's audit log answers that question - with full context for every action, every decision, every agent.

Why AI agents need audit trails

Traditional logging captures HTTP status codes and response times. Agent audit trails need to capture intent - what the agent tried to do, what policy evaluated it, and what decision was made. Without this, debugging agent behavior and proving compliance is impossible.

Compliance

GDPR, NIS2, and SOC 2 require audit trails for automated systems that process data or make decisions.

Incident response

When something goes wrong, you need to trace exactly which agent, which action, which policy, and which decision led to the outcome.

Debugging

Non-deterministic agents make different API calls each run. The audit trail shows exactly what happened on this specific run.

What TameFlare logs for every action

FieldExample
Agent (gateway)research-agent
Connectorgithub
Action typegithub.branch.delete
Parameters{ repo: "acme/backend", branch: "main" }
Decisiondenied
ReasonBranch deletion blocked by policy "protect-main"
Risk levelcritical
Timestamp2026-02-12T03:14:22Z
Latency12ms

The audit log is append-only. Events cannot be modified or deleted. Retention: 30 days (Starter), 90 days (Pro), 1 year (Team).

Audit log dashboard features

Full-text search

Search across agent names, action types, parameters, and reasons. Find any event instantly.

Filterable by type

Filter by event type, agent, connector, decision, or time range. URL-based filters for sharing.

CSV export

Export filtered audit data for compliance reporting, external analysis, or archival.

Action detail view

Click any event to see the full investigation view: action spec, policy trace, approval flow, and execution result.

Start logging agent actions today

Free Starter tier includes 30-day audit retention. No credit card required.