ce-babysit-pr

Watches an open GitHub PR and drives review feedback, CI failures, and branch currency toward merge-ready.

24.7k|2.0k|Updated Oct 9, 2025
One-click install
npx skills add https://github.com/EveryInc/compound-engineering-plugin --skill ce-babysit-pr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ce-babysit-pr
Source: https://github.com/EveryInc/compound-engineering-plugin/tree/main/skills/ce-babysit-pr
Command: npx skills add https://github.com/EveryInc/compound-engineering-plugin --skill ce-babysit-pr

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Open pull requests stall while authors juggle incoming review comments, failing CI checks, and base-branch drift across long review cycles. This Skill keeps a PR moving toward merge by continuously reacting to each of those streams as it arrives, instead of forcing you to poll and fix things manually.

Core Features & Use Cases

  • Continuous PR watch: Runs a background snapshot-based watcher that wakes only when review threads, CI failures, or branch-currency items need attention, with checkpoint and pipeline modes for harnesses without background wake support.
  • Delegated remediation: Routes review feedback to ce-resolve-pr-feedback and CI failures to ce-debug, handling comments before CI so fixes collapse both timelines instead of serializing them.
  • Managed stack support: Detects confirmed gh-stack managed stacks and supports target, stack-ready, and stack-land postures, including upstack propagation and authorized prefix landing.
  • Use Case: You open a PR before lunch and ask the agent to babysit it; hours later it has resolved reviewer threads, fixed a broken test, updated the branch when it fell behind, and reports it looks merge-ready with a full run recap.

Quick Start

Ask the agent to babysit the pull request for the current branch until it looks merge-ready.

Frequently Asked Questions about ce-babysit-pr

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I watch a GitHub pull request until it is merge-ready?

Invoke the skill with a PR number, URL, or the current branch and it starts a snapshot-driven watch loop. It resolves review threads, fixes failing CI, and maintains branch currency until the PR settles, is blocked, or hits its time budget.

How does it handle review comments and CI failures together?

It always processes review feedback before CI, delegating comments to ce-resolve-pr-feedback and failures to ce-debug. If the comment pass pushes a new commit, stale CI results on the old SHA are skipped instead of acted on.

Does it work with GitHub Enterprise or stacked pull requests?

Yes, it supports GitHub Enterprise hosts by keying state per host, and it detects confirmed gh-stack managed stacks. Postures target, stack-ready, and stack-land control whether it watches one PR, readies the whole stack, or lands settled prefixes.

Will it merge my pull request automatically?

No, merge-readiness is never merge authorization under the default target or stack-ready postures; the final merge stays with you. Only explicitly selecting the stack-land posture authorizes an automated gh stack merge of the settled prefix.

What happens when something needs a human decision?

Undecidable items become typed needs-human residuals that block the merge-ready declaration but do not stop the watch. The loop keeps driving independent review and CI streams and surfaces the full decision payload in its report.

Why does it stop working on GitLab or Bitbucket repositories?

The skill is GitHub-only because it and its delegates speak GitHub's API through gh, review threads, and Actions. If gh repo view fails or the remote points at GitLab or Bitbucket, it stops and reports that other forges are unsupported.