Build better skills through testing and iteration.
skill development
Creates, improves, evaluates, benchmarks, and optimizes reusable skills.
When to use it
Use when creating, editing, evaluating, benchmarking, or optimizing a skill.
Tell it what skill you want; it interviews you, drafts or edits the skill, tests it, gathers feedback, and iterates.
This skill
Claude
Sends trigger tests to Claude
documentation and similar skills
Researches docs and similar skills
generate_review.py
Starts a local review viewer
scripts.run_loop
Optimization in background
Python is needed for benchmark aggregation and other bundled scripts used by the evaluation workflow.
The claude CLI, specifically `claude -p`, is required only for description optimization.