pr-intel

Generate structured GitHub pull request review briefings with draft comments.

1|Updated May 15, 2026
One-click install
npx skills add https://github.com/mslshao/claude-harness --skill pr-intel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pr-intel
Source: https://github.com/mslshao/claude-harness/tree/main/dotclaude/skills/pr-intel
Command: npx skills add https://github.com/mslshao/claude-harness --skill pr-intel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you review pull requests faster and with better judgment by turning a raw PR into a structured intelligence briefing. It reduces the risk of missed regressions, duplicated comments, shallow reviews, and unclear verdicts by gathering context, checking prior review history, analyzing code changes, and drafting comments you can post directly.

Core Features & Use Cases

  • Structured PR review briefings: Builds a reviewer-facing summary with scope, recommendation, verdict, and draft inline comments.
  • Context-aware analysis: Pulls PR metadata, diffs, prior reviews, CI status, Jira acceptance criteria, design docs, service context, and bot comments.
  • Specialist-backed review: Dispatches focused analysis for security, tests, infrastructure, TypeScript, observability, history, blast radius, configuration, and skeptical edge-case questioning.
  • Dedup and verification workflows: Avoids repeating prior comments, classifies findings by provenance, reacts to bot feedback, and scales verification depth by PR size.
  • Use cases: Use it when reviewing a teammate's PR, doing a pre-submit self-review with --mine, triaging a PR quickly with --quick, or preparing a full briefing before posting GitHub review comments.

Quick Start

Ask the AI to review a pull request by number or URL, for example by saying review PR 123 with pr-intel and focus on risky changes around tests, security, or acceptance criteria.

Frequently Asked Questions about pr-intel

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

FAQPage Schema
How do I automate GitHub pull request review for code quality and security?

Automate GitHub pull request review by dispatching specialist-backed analysis for code quality, security, and infrastructure to produce structured briefings and draft inline comments. This workflow hydrates context from diffs and CI status to generate reviewer-ready intelligence.

Can I check Jira acceptance criteria against a pull request diff automatically?

Yes, you can check Jira acceptance criteria against a pull request diff by hydrating PR context with Jira metadata and design docs. The review intelligence verifies if the code changes fulfill the required project acceptance criteria before posting a verdict.

What is the best way to do a self-review of my PR before submitting it?

The best way to self-review a PR is using a one-shot analysis mode that checks your own diff for regressions, security holes, and test coverage. It produces a structured briefing highlighting risky changes and draft comments to fix issues before requesting teammates.

How do I avoid posting duplicate review comments on a GitHub pull request?

Avoid posting duplicate review comments by analyzing prior review history and bot feedback before generating new findings. The deduplication workflow classifies findings by provenance and ensures your drafted GitHub comments add new value rather than repeating existing feedback.

Does pull request analysis work for quick triage of large changes?

Yes, pull request analysis works for quick triage by scaling verification depth based on PR size. A quick-triage mode focuses on blast radius, CI status, and critical security risks to provide an immediate verdict without deep-diving into every minor change.

What context is needed for automated PR analysis to generate review comments?

Automated PR analysis requires PR metadata retrieval, diff analysis, and context hydration from service context, CI status, and prior reviews. This contextual data fuels the specialist dispatch process to verify findings and format output suitable for direct posting to GitHub.