メニュー
Run QA validation on a pull request — boots the local environment, tests acceptance criteria, and optionally posts the report as a PR comment. Standalone entry point for the qa-engineer agent.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | maestro:qa |
| description | Run QA validation on a pull request — boots the local environment, tests acceptance criteria, and optionally posts the report as a PR comment. Standalone entry point for the qa-engineer agent. |
| argument-hint | <PR-number-or-URL> |
Standalone QA run for any PR. Boots the local environment, validates every acceptance criterion, and produces a test report. Posting to GitHub is your choice — you are prompted at the end.
Use $ARGUMENTS as the PR number or URL. If empty, resolve from the current branch:
gh pr list --head "$(git branch --show-current)" --json number,url -q '.[0] | "\(.number) \(.url)"'
If no PR is found, tell the user and stop.
Read .claude/maestro.json. Extract:
REPO = .ai.repoTEMP_ROOT = .ai.temp_rootInvoke the qa-engineer sub-agent with:
gh pr view <PR_NUMBER> --json baseRefName -q .baseRefName)STANDALONE MODE — two differences from the normal pipeline run:
- Skip Step 6 (posting the PR comment). Instead, output the full QA report as formatted Markdown in your response, in a section titled
## QA Report. Use the same format the pipeline would post (including the<!-- ai-pipeline:qa-report -->marker) so it is ready to post as-is.- Skip the StructuredOutput JSON return. Output a short human-readable summary instead: overall result, pass/fail per criterion, and any blockers.
All other steps run normally — the environment is booted, acceptance criteria are tested, and the full validation is performed. The report is always complete regardless of posting.
After the agent responds, display its ## QA Report and ask:
Post this QA report to PR #<PR_NUMBER>? Reply
yesto post,noto finish here.
If yes — post using the dedup flow: check for an existing <!-- ai-pipeline:qa-report -->
comment, update it with PATCH if found, otherwise create a new comment. This means a
subsequent pipeline re-run will update the comment in place rather than duplicate it.
If no — confirm the QA run is complete and finish.
Generate a bespoke issue-workflow for any target project. Phase 1 — a Claude Opus analyst reads the project deeply and produces a disposition report. Phase 2 — on approval, parallel writer agents transplant every Maestro workflow component, adapted to the target project's actual stack, test runner, dev environment, and conventions. Use when a project needs its own self-contained issue-workflow without taking a direct dependency on Maestro.
Show the Maestro command map — a quick reference to all available skills.
Adversarially review a grooming spec before implementation starts. Finds hidden risks, unvalidated assumptions, and missing dependencies. Standalone entry point for the challenger agent.
Run a lead code review on the current branch or a given PR. Standalone entry point for the lead-reviewer agent.
Groom a single GitHub issue — produce an implementation spec and optionally post the grooming summary as a GitHub comment. Standalone entry point for the grooming-agent.
Check a change against WordPress.org plugin rules and PHPCS standards.