skill-freshness-audit

To check if MLflow agent skills are outdated, perform a systematic freshness audit to verify skill content against live official documentation and track upstream lineage. This process identifies staleness and reports drift to ensure your skills match current APIs.

Upstream lineage tracking for AI dev kit skills is the process of monitoring changes in dependent repositories to detect when updates occur. It classifies skill volatility and reports drift to prevent implementation errors from outdated information.

To detect staleness in Databricks agent skills before a platform release, run a freshness audit that checks skills against live documentation. It identifies skills that haven't been verified recently based on their classified volatility levels.

The best way to automate documentation verification for agent skills is performing a systematic skill freshness audit. This approach regularly checks skills against live documentation sources, identifies staleness, and reports drift across frameworks like Databricks and MLflow.

The skill freshness audit works with custom AI dev kit repositories by tracking upstream lineage and monitoring changes in dependent repositories. It systematically verifies skills against official documentation to identify staleness and report drift.