What problem does it solve? Managing a growing library of Claude Skills leads to version drift, lost updates from parallel sessions, unvalidated changes, and broken packages. This Skill enforces a governed lifecycle for every skill: re-read from disk before editing, show diffs before applying, validate with a regression guard, bump versions, and package safely. ## Core Features & Use Cases - Master Administrator Protocol: A command table (update skill, add rule, version bump, repackage, rollback) with a mandatory diff-and-confirm workflow before any change is applied. - Three Laws Safety Core: Immutable priority hierarchy that blocks harmful directives such as deleting safety rules or skipping validation. - Self-Development Engine: Post-use gap and pattern assessment that generates improvement proposals without applying them autonomously. - Regression Guard & Packaging: A Python guard script validates canonical terms and line counts, and packaging rules produce upload-safe .skill files. - Use Case: You ask to update a skill's routing rules; the Skill re-reads the current file from disk, drafts the change, shows you a diff, waits for approval, then updates the CHANGELOG, bumps the version, snapshots, and repackages. ## Quick Start Ask the assistant to update a named skill, for example by saying it should add a new rule to the semantic-router skill and bump its version.
