If you're building or maintaining Claude skills, this is your workshop. It handles the full authoring cycle: drafting SKILL.md files, structuring folders (references, scripts, assets), ingesting external docs, running trigger evals, and benchmarking quality through blind A/B tests. It follows the agentskills.io spec and includes templates, validation scripts, and an eval review UI. The iterative improvement workflow is particularly thorough, with workspace layout guides and benchmark analysis. Use it when you need to create skills from scratch, optimize existing descriptions, or validate before publishing. It replaces the older skill-creator approach with a more structured operator playbook.
npx skills add https://github.com/itechmeat/llm-code --skill skill-master