What problem does it solve? Maintaining a large skill library makes it hard to know which skills are actually used, whether cron-scheduled tasks still run, and which directories are orphaned. This Skill produces auditable skill usage statistics and prevents accidental deletion of directories that are still referenced by active skills. ## Core Features & Use Cases - Sourced Usage Reporting: Scans the skills/ directory and cron task list, then generates usage reports (skill counts, cron task counts, total run counts, archived tasks) where every number carries a traceable collection command. - Consistent v1/v2 Data: Keeps report versions on the same measurement basis, using evolution-state.json as the single data source. - Safe Unmapped Directory Cleanup: Detects directories with no cron references and no invocation records, verifies they are not referenced by live skills before cleanup, and rolls back instead of deleting anything still in use. - Use Case: A repository with 27 skills and 21 cron tasks needs a usage audit. The Skill produces a v2 usage report showing 1352 total runs and 3 archived tasks, and flags skills/old-experiment/ as a candidate for cleanup only after confirming no live skill references it. ## Quick Start Scan the skills directory and cron tasks, then generate a sourced skill usage report and list any unmapped directories safe to clean.
