From AI Dev Pipeline
Use when asked to improve the AIDP plugin itself by finding exactly one small maintainer-reviewable bug, docs ambiguity, confusing error, test gap, or contributor-experience papercut.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ai-dev-pipeline:aidp-maintainer-improvementThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill to turn agent attention into one reviewable contribution. Do not run an autonomous loop, do not keep searching after a patch is ready, and do not broaden into product direction.
Use this skill to turn agent attention into one reviewable contribution. Do not run an autonomous loop, do not keep searching after a patch is ready, and do not broaden into product direction.
Start a typed AIDP run:
aidp "find one small maintainer-reviewable improvement" --run-id <short-id> --proof-command "uv run pytest -q"
Read .aidp/runs/<run-id>/work.md, then inspect only the repo surfaces needed to choose one target.
Choose exactly one of:
Prefer a failing test or docs reproduction over a broad refactor.
aidp check <run-id>, aidp prove <run-id>, and aidp finish <run-id> before claiming done.Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub 0x7067/ai-dev-pipeline --plugin ai-dev-pipeline