概要
This specialized capability provides a rigorous framework for maintaining high-quality Claude skills through a systematic 9-phase auditing process. It effectively bridges the gap between automated scripts and human-like reasoning by verifying code patterns against official documentation, detecting version drift in dependencies, and identifying cross-file inconsistencies. Whether preparing for a marketplace submission or updating a skill after a major package release, this tool ensures that instructions are accurate, examples are functional, and performance is optimized for maximum token efficiency.