audit

You can audit conversation context against governance principles by invoking a one-shot audit command that triggers a stronger advisor model to compare your active context against curated principles and return a structured findings report.

Stress-testing decisions non-interactively involves running an automated audit that identifies gaps between your current context and curated principles, generating an Audit Summary table and logging usage to a tracker file.

Yes, when auditing context traceability you can select principles automatically by default or specify explicit principle IDs to target exactly which governance rules the advisor model evaluates against your active context.

Logging traceability audits is handled automatically; the audit process writes usage data to a tracker markdown file and returns a structured audit output separating content verdicts from execution results for downstream consumption.

The audit advisor supports automatic principle selection, identifying relevant governance rules for your context without manual input, while also allowing explicit specification of principle IDs for targeted reviews.

The structured audit output format includes an Audit Summary table presenting findings clearly, alongside an audit log, with content verdicts separated from execution results and graceful error handling built in.