Your prompts need tests too. Run prompts against real datasets, score outputs with LLM judges, version everything, and compare runs to see what got better.
Latest release v0.28.42 · 12 Aug 2026
2 files for Claude Code: contributor-pr-assessor, completion-kit CLAUDE.md — 341 tokens loaded in every session.
CLAUDE.md A 341 tok .claude/agents/contributor-pr-assessor.md A 90 tok These files are homemade-software-inc/completion-kit's own configuration — they tell Claude Code how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.