iterate-pr
Iterate on pull requests until CI passes
๋ํ ๋ค์์์ ์ฌ์ฉํ ์ ์์ต๋๋ค: sickn33,getsentry
Managing CI failures and review feedback requires repeatedly checking status, applying fixes, and pushing changes. This skill automates the feedback-fix-push cycle so developers can focus on writing code instead of manually tracking CI pipelines.
์คํฌ ZIP ๋ค์ด๋ก๋
Claude์์ ์ ๋ก๋
์ค์ โ ๊ธฐ๋ฅ โ ์คํฌ โ ์คํฌ ์ ๋ก๋๋ก ์ด๋
ํ ๊ธ์ ์ผ๊ณ ์ฌ์ฉ ์์
ํ ์คํธํด ๋ณด๊ธฐ
"iterate-pr" ์ฌ์ฉ ์ค์ ๋๋ค. My PR has 2 failing tests and a review comment. Fix them.
์์ ๊ฒฐ๊ณผ:
- Found PR #123 for branch feature/login
- CI Status: 2 failed (test-unit, test-integration)
- Found 1 review comment from @reviewer
- Analyzing test failures...
- Addressing review comment...
- Changes committed and pushed
- Watching CI... (checks passing: 5/7)
- Waiting for remaining checks...
- All CI checks passed! PR ready for merge
๋ณด์ ๊ฐ์ฌ
๋ฎ์ ์ํDocumentation-only skill providing CI/CD workflow guidance. Uses standard GitHub CLI commands for PR iteration. All commands match the stated purpose of PR automation. Static findings are false positives - backticks flagged as shell execution are actually markdown code block markers in documentation.
๋ฎ์ ์ํ ๋ฌธ์ (1)
์ํ ์์ธ
โ๏ธ ์ธ๋ถ ๋ช ๋ น์ด (4)
๐ ํ์ผ ์์คํ ์ก์ธ์ค (1)
ํ์ง ์ ์
๋ง๋ค ์ ์๋ ๊ฒ
Fix CI failures automatically
Continuously fix test failures and linting issues until all CI checks pass without manual polling
Address review feedback efficiently
Track and resolve human review comments across multiple iteration cycles until approval
Monitor bot feedback
Wait for Sentry, Codecov, and other bot checks to complete before addressing their feedback
์ด ํ๋กฌํํธ๋ฅผ ์ฌ์ฉํด ๋ณด์ธ์
Use the iterate-pr skill to fix the failing CI checks on my current PR
Check my PR for review feedback and address all comments using iterate-pr
Run iterate-pr to ensure all CI checks pass and all feedback is addressed before merge
Use iterate-pr to investigate why my CI is failing repeatedly and determine if it needs user intervention
๋ชจ๋ฒ ์ฌ๋ก
- Verify the PR exists for your current branch before starting the iteration
- Wait for bot-related checks (Sentry, Codecov) to complete before addressing their feedback
- Validate that feedback is legitimate and not already addressed before making changes
- Let the skill stop and ask for help when failures persist beyond 3 attempts
ํผํ๊ธฐ
- Running the skill without GitHub CLI authentication or proper repository permissions
- Ignoring pending CI checks and addressing feedback too early
- Assuming all feedback is correct without validating the issue first
- Continuing iteration when CI failures are infrastructure issues unrelated to your code
์์ฃผ ๋ฌป๋ ์ง๋ฌธ
Does this skill work with GitHub Enterprise?
How long does it wait between CI checks?
Can I customize which CI checks to wait for?
Is my data safe when using this skill?
What happens if CI keeps failing?
How is this different from GitHub Actions auto-merge?
๊ฐ๋ฐ์ ์ธ๋ถ ์ ๋ณด
์์ฑ์
davila7๋ผ์ด์ ์ค
MIT
๋ฆฌํฌ์งํ ๋ฆฌ
https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/sentry/iterate-pr์ฐธ์กฐ
main
ํ์ผ ๊ตฌ์กฐ
๐ SKILL.md