Skill Expansion Packs

•
35
•

Manages third-party Claude Code skill imports, forks, and lifecycle states with clean namespace isolation and provenance tracking.

The Skill Expansion Packs skill provides a standardized package management workflow for importing, organizing, and maintaining third-party Claude Code skills and prompt stacks. It enforces explicit namespacing and directory layouts to prevent registry collisions, allows developers to safely fork community tools into tracked mutations, and supports beta sandboxing with manual-only execution flags. This skill is indispensable for developers managing large collections of AI agent capabilities who need clear provenance, easy diffing against upstream repositories, and seamless integration into central skill dashboards.

主な機能

01Isolated mutation workflows (grim:ep:mx:) to fork and customize community tools
0235 GitHub stars
03Safe beta lifecycles with default manual-only execution to prevent unintended invocations
04Declarative pack manifest system (pack.yaml) for managing multi-skill stacks
05Strict namespace isolation separating native, third-party, and mutated skills
06Structured upstream diffing workflows to preserve local modifications during updates

ユースケース

01Sandboxing experimental or beta agent capabilities before promoting them to active workflows
02Importing third-party agent skills and prompts without namespace collisions
03Customizing community Claude Code tools while tracking upstream diffs and license provenance