ワンクリックで
monitor-pr
Monitor the CI checks status of a Pull Request on GitHub, parse logs on failure, and address any non-passing tests.
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。
メニュー
Monitor the CI checks status of a Pull Request on GitHub, parse logs on failure, and address any non-passing tests.
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。
Start work on a new GitHub issue. Fetches issue details, generates standard branch names, and sets up the branch.
Generate a detailed, high-depth pull request body and create/update a PR on GitHub. Gathers workspace changes, updates the PR body dynamically, and posts update comments to capture a historical paper trail.
Raise a detailed, structured issue on GitHub. Gathers context, builds reproduction steps, expected outcomes, and acceptance criteria.
Finish work on the current issue branch. Runs quality gates (formatting, linting, tests), handles uv.lock version drift, and pushes to remote.
SOC 職業分類に基づく
| name | monitor-pr |
| description | Monitor the CI checks status of a Pull Request on GitHub, parse logs on failure, and address any non-passing tests. |
Standardized procedure to track pull request CI checks on GitHub and address any test or linting failures.
raise-pr), or when the user requests to check, monitor, or track the CI/test status of an active PR.Check CI checks status:
Use the GitHub CLI (gh) to retrieve the status of all active checks on the current branch:
gh pr checks
Wait for completion (if pending):
If any checks are marked as pending:
schedule tool (e.g., set for 120 seconds with the prompt "Check CI status for PR").sleep loop or poll continuously in a terminal process.Handle successful runs:
If all checks are marked as pass or success:
Diagnose and fix failures (if any check fails):
If any check fails (status is fail or failure):
test, lint, docs-lint).gh run list --branch <current-branch> --limit 1
gh run view <run-id> --log-failed
make test (or run a specific test: uv run pytest -k <failing_test_name>)make lint or make formatmake format && make lint && make test) to confirm the fix.fix: address failing CI check), and push them to origin (git push).