drive-pr

Resolve trusted GitHub PR feedback, fix CI failures, and update descriptions.

Updated Jun 3, 2026
One-click install
npx skills add https://github.com/0xdeafcafe/skills --skill drive-pr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: drive-pr
Source: https://github.com/0xdeafcafe/skills/tree/main/skills/drive-pr
Command: npx skills add https://github.com/0xdeafcafe/skills --skill drive-pr

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill takes over the repetitive work of finishing an active pull request by resolving trusted review feedback, fixing failing required CI checks, and keeping the PR description aligned with the actual code. It is designed for the last-mile coordination work that often slows shipping even after the core implementation is done.

Core Features & Use Cases

  • Trusted review resolution: Fetches PR comments, review threads, and reviews, then acts only on feedback from verified humans or explicitly whitelisted bots.
  • Parallel code-fix workflow: Converts actionable review comments into structured findings, batches them through an agent pipeline, and applies safe edits in parallel before replying and resolving threads.
  • CI and PR body completion: Checks required status checks, fixes failing runs when possible, and updates the PR description so it accurately matches the shipped diff and template requirements.
  • Use case: After opening a GitHub pull request, use this Skill to chase down CodeRabbit comments, human reviewer fixes, red required checks, and stale PR text until the branch is ready to merge.

Quick Start

Ask the agent to use drive-pr on the current branch's pull request and keep iterating until trusted comments are resolved, required CI is green, and the PR description matches the code.

Frequently Asked Questions about drive-pr

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

FAQPage Schema
How do I automate resolving GitHub pull request comments and CI failures?

Automate pull request resolution by fetching trusted review feedback, fixing failing required CI checks, and updating the PR description. This skill verifies commenter trust policies and applies batched code edits through coordinated subagent workflows until the GitHub PR is merge-ready.

Can I use this to fix required CI checks and CodeRabbit bot feedback on a GitHub PR?

Yes, you can use this to fix required CI check regressions and resolve CodeRabbit bot feedback. It handles automated bot feedback, human reviewer comments, and post-push CI failures, applying safe edits in parallel before replying to and resolving GraphQL review threads.

What's the best way to drive a pull request to merge-ready status automatically?

The best way to drive a pull request to merge-ready status is to iterate on trusted review comments, fix required CI failures, and align the PR description with the shipped diff. It converts actionable comments into structured findings and applies safe edits in parallel.

Do I need GitHub CLI access to automate pull request review resolution?

Yes, you need GitHub CLI access. The pull request review resolution process requires GitHub CLI access to fetch PR comments, verify trust policies for commenters, synchronize git branches, and perform GraphQL thread resolution.

How does trust verification work when automating GitHub PR code reviews?

Trust verification for GitHub PR code reviews works by checking that feedback comes from verified humans or explicitly whitelisted bots. It fetches PR comments and reviews, then acts only on trusted feedback to safely apply code changes and resolve threads.

Why are not all pull request review comments automatically resolved when using review automation?

Not all pull request review comments are resolved because the automation only acts on feedback from verified humans or explicitly whitelisted bots. This trust policy prevents applying untrusted code changes while safely batching and applying actionable structured findings.