skill-lifecycle

Automated archiving of unused skills is handled by running a lifecycle sweep command that evaluates usage telemetry and moves low-value directories to an archive. This process is reversible and takes pre-archive snapshots to ensure skills remain restorable.

Skill usage telemetry tracks views, usage, and patches for every skill to calculate its actual value. This tracked activity data is then used by lifecycle scripts to enforce value-based retention policies and identify idle skills for archiving.

Python 3 is required to execute lifecycle scripts and maintain the .usage.json sidecar and .lifecycle-journal.jsonl log. These scripts track activity, enforce retention policies, and maintain an audit journal of all skill movements.

You can protect specific skills from auto-archiving by pinning them to exempt them from lifecycle transitions. The system also includes built-in safety guards that automatically protect core and bundled skills from being archived.

Maintaining an audit journal of skill state transitions is achieved by logging all movements to a .lifecycle-journal.jsonl file. The lifecycle management system records every auto-archiving and state transition event in this reversible log.

Stale skill accumulation is prevented by a structured lifecycle management system that uses witnessed telemetry and auto-archiving. It ensures no work is permanently deleted by moving idle skills to an archive with pre-archive snapshots.