governance-trace-audit

To audit Langfuse traces for AI governance, you can run this Skill to analyze trace data from the AgentsFramework runtime and validate it against governance rules. It assesses traces across the governance triangle to ensure compliant agent behavior.

The governance triangle in AI agent trace audits consists of Recording, Identity, Validation, and Reasoning pillars. This Skill validates agent behavior against these specific governance criteria to ensure operational traceability and compliance.

You validate agent behavior after a deployment by running an audit on the latest runtime traces. This Skill compares trace data against governance criteria and generates a detailed report with scores for each pillar and an overall compliance status.

Yes, you can check trace compliance without custom scripts by running this Skill. It directly processes Langfuse trace data from the AgentsFramework runtime, automatically applying governance rules and outputting a detailed compliance verdict.

Yes, the governance audit generates a detailed compliance report that includes specific scores for each governance pillar—Recording, Identity, Validation, and Reasoning—along with an overall compliance status verdict for the audited agent runtime traces.

You should run a trace audit on your AI agent runtime after deploying any changes to the agent. This ensures that the new runtime behavior remains fully traceable and adheres to predefined governance rules before you proceed further.