Draft SKILL.md content and run trigger/evaluation loops for Claude Code skills.
npx skills add https://github.com/junevm/avtoolz --skill skill-creator-junevm
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/junevm/avtoolz/tree/main/.agents/skills/skill-creator Command: npx skills add https://github.com/junevm/avtoolz --skill skill-creator-junevm
Skill-creator helps you reliably create and iterate on Claude skills by turning your goals into a drafted SKILL.md, generating realistic test prompts, running trigger/evaluation checks, and using results to refine the skill for better triggering accuracy and performance.
Use skill-creator when you want to build a new Claude skill for “processing browser-side PDFs” from scratch, then iteratively refine its description and eval results until it triggers at the right moments.
High-intent search queries and answers about installing and using this skill.
To create and test a Claude Code skill, draft a SKILL.md file defining your intent, generate test prompts, and run trigger/evaluation loops. This iterative process refines the skill's triggering accuracy and task performance.
Your Claude skill triggers incorrectly due to an unoptimized YAML frontmatter description. Improve triggering accuracy by running benchmark evaluations on should-trigger and should-not-trigger queries, then iteratively rewriting the description.
The best way to evaluate and benchmark Claude skills is to build a test set, run evaluations against a baseline, and use grader aggregation. This quantifies behavioral impact and produces review artifacts for qualitative decisions.
Yes, you need the Claude dependency to automate skill design and trigger optimization. It executes the root SKILL.md-based folder workflow, generates realistic test prompts, and performs the evaluation runs required for iterative revision.
Yes, you can improve an existing Claude skill without a full rebuild. Apply targeted description optimization and run trigger/evaluation checks to refine the YAML frontmatter, increasing correct activation and reducing false triggers.
Menschlichkeit-OsterreichCreate, test, and optimize AI skills with automated evaluations.
niduran-orionCreate, evaluate, and refine Claude skills with SKILL.md and trigger evaluations.
zhmxiaowoDefine automated evaluation criteria for Claude Code tasks with multiple graders.
pipixia-labsGuide creation of SKILL.md files and coordinate evals, benchmarks, and iteration.
