babysitting-pr — Claude Code skill: what it does, how to install
Monitors an open pull request for CI failures, review comments, and merge conflicts, then fixes issues locally and pushes updates.
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a claude-codeWhat babysitting-pr does
This skill keeps an open pull request ready for merging. The agent checks PR status through the GitHub CLI, pulls logs for failing CI checks, and inspects errors locally. It handles linter problems, TypeScript typing bugs, and broken builds before pushing a commit back to the branch.
When merge conflicts occur, the agent merges origin/main to resolve them. It also fetches unresolved review comments. Straightforward requests like typos or null checks get resolved right away. For ambiguous conflicts or architectural choices, the agent stops and asks the user.
Execution relies entirely on an authenticated gh CLI tool. The process runs through a verify-and-push loop that stops after 3 attempts. This prevents runaway loops on flaky test suites.
What it helps with
- fix failing CI lint and build checks
- resolve merge conflicts against origin main
- address straightforward review comments automatically
- keep pull requests merge-ready without manual monitoring
Example prompts for the agent
- Check the open PR and fix failing CI checks
- Resolve merge conflicts with main and address review comments
- Monitor the current pull request until checks are green
How to install babysitting-pr
1Skills CLI (Vercel)
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a claude-code2Manually: put the files in place
~/.claude/skills/babysitting-pr/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/babysitting-pr/
1Skills CLI (Vercel)
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a codex2Manually: put the files in place
.agents/skills/babysitting-pr/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a cursor2Manually: put the files in place
.cursor/skills/babysitting-pr/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install spencerpauly/awesome-cursor-skills babysitting-pr2Manually: put the files in place
.github/skills/babysitting-pr/SKILL.md1Skills CLI (Vercel)
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a gemini-cli2Platform CLI
gemini skills install https://github.com/spencerpauly/awesome-cursor-skills --path resources/babysitting-prManually: put the files in place
.devin/skills/babysitting-pr/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about babysitting-pr
How to install babysitting-pr in Claude Code?
Run npx skills add spencerpauly/awesome-cursor-skills --skill babysitting-pr -a claude-code in your terminal.
What does babysitting-pr do?
It checks pull request health via GitHub CLI, fixes CI failures and simple review comments, and resolves merge conflicts in a repeating loop.
When does the PR monitoring loop stop?
It stops when all checks pass, when an issue requires a design decision, or after 3 unsuccessful fix-and-check cycles.
Source text
SKILL.md · CC0-1.0
---
name: babysitting-pr
description: Monitor a pull request for CI failures, review comments, and merge conflicts — then fix them automatically. Use when a PR is open and you want the agent to keep it merge-ready.
---
# Babysitting a PR
Use this skill when the user has an open pull request and wants the agent to monitor it, fix CI failures, resolve review comments, and keep it merge-ready.
## Steps
1. **Get the PR status** — fetch the current state of the PR:
```bash
gh pr view --json number,title,state,mergeable,reviewDecision,statusCheckRollup,comments,reviews
```
Also check for merge conflicts:
```bash
gh pr view --json mergeStateStatus
```
2. **Check CI status** — look at the `statusCheckRollup` field. For each failing check:
```bash
gh pr checks
```
This lists all CI checks and their status (pass/fail/pending).
3. **Fix CI failures** — for each failing check, get the logs:
```bash
gh run view <run-id> --log-failed
```
Analyze the failure and fix it:
- **Lint failures**: run the linter locally (`npm run lint -- --fix`), fix remaining issues manually, commit.
- **Type errors**: run `npx tsc --noEmit`, read the errors, fix the types, commit.
- **Test failures**: run the failing test suite locally, read the assertion errors, fix the code or update the test expectations, commit.
- **Build failures**: run `npm run build`, read the error output, fix imports/configs/missing deps, commit.
After fixing, push the changes:
```bash
git add -A && git commit -m "fix: resolve CI failures" && git push
```
4. **Handle review comments** — fetch PR review comments:
```bash
gh api repos/{owner}/{repo}/pulls/{pr}/comments
```
For each unresolved comment:
- Read the comment and the code it references.
- If the fix is clear (typo, naming, missing null check, style issue), apply it.
- If the comment requires a design decision or clarification, skip it and report to the user.
Commit fixes for resolved comments:
```bash
git add -A && git commit -m "fix: address review feedback" && git push
```
5. **Resolve merge conflicts** — if the PR has conflicts:
```bash
git fetch origin main
git merge origin/main
```
Resolve conflicts by reading both sides and choosing the correct resolution. For ambiguous conflicts, ask the user. After resolving:
```bash
git add -A && git commit -m "fix: resolve merge conflicts" && git push
```
6. **Re-check status** — after pushing fixes, wait for CI to run and check again:
```bash
gh pr checks --watch
```
If new failures appear, go back to step 3. Limit to 3 rounds to avoid infinite loops.
7. **Report** — summarize what was done:
- Which CI checks were failing and how they were fixed
- Which review comments were addressed
- Whether merge conflicts were resolved
- Current PR status (ready to merge, or what's still blocking)
## Loop Behavior
This skill is designed to run in a loop:
```
Check PR → Find issues → Fix issues → Push → Re-check → Repeat
```
Stop when:
- All checks pass, no unresolved comments, no conflicts → PR is merge-ready
- 3 fix-push-check cycles have been attempted without full resolution → report what's still failing
- A fix requires a design decision → ask the user
## Notes
- Never force-push to a shared PR branch.
- Don't modify test assertions to make tests pass unless the behavior change was intentional.
- Don't resolve review comments you're unsure about — skip them and let the user know.
- If CI is queued/pending, wait for it to complete before analyzing failures.
- Use `gh pr ready` to mark the PR as ready for review once everything is green.
© spencerpauly · CC0-1.0 · spencerpauly/awesome-cursor-skills