github-review

Automate GitHub PR code reviews with AI-generated inline comments and summary artifacts.

132|13|Updated Dec 18, 2025
One-click install
npx skills add https://github.com/holon-run/holon --skill github-review-holon-run
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: github-review
Source: https://github.com/holon-run/holon/tree/main/skills/github-review
Command: npx skills add https://github.com/holon-run/holon --skill github-review-holon-run

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Automate the end-to-end PR review process by collecting context, running AI-based analysis, and publishing structured, actionable feedback as inline comments and summaries.

Core Features & Use Cases

  • AI-powered code review of changes in GitHub PRs with inline comments and a consolidated summary.
  • Context collection: gathers PR metadata, diffs, changed files, and history to inform the review.
  • Publish and track feedback: posts inline comments on the PR and outputs artifacts like review.md, review.json, summary.md, and manifest.json for CI and auditing.
  • Use Case: When a PR opens or updates, run a headless review to surface correctness, security, and quality findings and attach them to the PR.

Quick Start

Trigger the review by providing a PR reference (owner/repo#PR) to Holon to collect context, analyze changes, and publish findings.

Frequently Asked Questions about github-review

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

FAQPage Schema
How do I automate GitHub pull request code reviews with inline comments?

Automate PR code reviews by providing a PR reference to collect context, analyze changes, and publish AI-generated findings as inline comments. It integrates with CI workflows or manual triggers to surface correctness, security, and quality feedback directly on the pull request.

What context does an AI code review gather to analyze a pull request?

AI code review gathers PR metadata, diffs, changed files, and history to inform the analysis. This collected context allows the system to produce structured, actionable feedback and consolidated summaries based on the actual changes in the pull request.

Can I use automated PR review in my existing CI/CD pipeline?

Yes, automated PR review works within CI/CD pipelines or manual workflows. It integrates with the gh CLI for authentication and tokens, posting inline comments on the PR and outputting artifacts like review.md, review.json, and summary.md for CI auditing.

What artifacts are generated when running an automated GitHub PR review?

Running an automated GitHub PR review generates review.md, review.json, summary.md, and manifest.json artifacts. These files are output to standard directories, providing structured records of the AI analysis and consolidated summary for CI and auditing purposes.

Do I need the gh CLI to publish inline comments on a GitHub pull request?

Yes, the gh CLI is required for authentication and token management to publish inline comments on a GitHub pull request. The review process satisfies integration with gh CLI and supports an optional ghx publishing flow to attach feedback directly to the PR.

What is the best way to track AI-generated findings from a headless PR review?

The best way to track AI-generated findings is through the outputted summary.md and review.json artifacts. These files consolidate the correctness, security, and quality findings, allowing you to audit the headless review results alongside the posted inline comments.