Which specialized agents are included in the review?

The review typically includes agents for Code Quality, Type Safety, Security, Test Coverage, Backend Architecture, and Frontend Development. An optional LLM Integrator agent is available for AI-related code.

What is the difference between Agent Teams and Task tool modes?

Agent Teams mode allows reviewers to cross-reference findings directly (mesh communication) for deeper analysis, while Task tool mode uses a centralized lead agent (star communication) for faster, lower-cost reviews.

Does this skill require the GitHub CLI?

Yes, the skill uses 'gh' commands to view PR metadata, fetch diffs, and check CI status directly from your local environment.

Can I focus the review on specific areas like security?

Yes, at the start of each review, the skill uses the AskUserQuestion tool to let you choose between a Full Review, Security Focus, Performance Focus, or a Quick Review.

Deep PR Reviewer

Name: Deep PR Reviewer
Author: yonatangross

byyonatangross

•

Collaboration Tools

Automates comprehensive pull request reviews using a swarm of specialized AI agents to analyze security, quality, and architecture.

The Deep PR Reviewer skill transforms the standard code review process by deploying up to seven parallel specialized agents—including security auditors, backend architects, and frontend developers—to provide multi-perspective feedback on any pull request. Built for the OrchestKit framework, it integrates directly with the GitHub CLI to fetch diffs, run validation scripts, and synthesize findings into a structured, production-ready report. Whether you need a quick quality check or a deep security audit, this skill manages the complex orchestration required to catch bugs and architectural flaws before they reach production.

Key Features

01Deep security auditing for secrets, injection, and vulnerabilities

02Specialized reviews for React 19, TypeScript, and Backend APIs

0369 GitHub stars

04Adaptive orchestration modes (Mesh vs. Star) for cost and depth control

05Parallel agent orchestration for multi-domain analysis

06Automated CI/CD status checks and validation script execution

Use Cases

01Standardizing code quality and type safety across large engineering teams

02Accelerating review cycles for complex PRs with cross-cutting architectural changes

03Performing rigorous security audits on critical production feature branches

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yonatangross/orchestkit review-pr

For use in Claude.ai and ChatGPT

Download Skill