Can the skill actually modify my code?

Yes, the pipeline is designed to automatically apply fixes for 'Critical' and 'High' severity findings to streamline your development process.

What happens if the Codex CLI is not installed?

The pipeline includes a pre-flight check; if Codex is unavailable, it gracefully falls back to a Claude-only review without interrupting the workflow.

Is it safe to use on large diffs?

The pipeline uses parallel agents with their own context windows, making it highly efficient for reviewing substantial changes across multiple files.

Which AI models does the pipeline use for reviews?

The pipeline primarily utilizes Claude Opus for agent teammates and can optionally integrate the Codex CLI to provide parallel cross-validation of findings.

How does it decide which reviewers to use?

The skill analyzes the file extensions in your git diff to dispatch specific reviewers for UI, Architecture, Tests, and Implementation logic automatically.

Code Review Pipeline

Name: Code Review Pipeline
Author: guyathomas

byguyathomas

0•

Collaboration Tools

Orchestrates a multi-agent team to perform parallel code reviews, cross-validate findings, and automatically resolve critical issues.

The Code Review Pipeline transforms your development workflow by deploying specialized agent teams—including architecture, UI, and test specialists—to analyze git diffs concurrently. By leveraging a dual-engine approach with Claude Opus and optional Codex background processes, the skill filters out low-confidence noise and highlights cross-validated vulnerabilities. It goes beyond simple reporting by automatically implementing fixes for high-severity issues and providing a detailed synthesis of architectural improvements and missing test coverage before you commit your code.

Key Features

01Dual-engine validation combining Claude Opus and Codex for high precision

02Automated remediation of critical and high-severity code issues

03Comprehensive reporting of missing test paths and architectural suggestions

04Parallel multi-agent review team tailored to specific file types

050 GitHub stars

06Intelligent finding synthesis with noise filtering and engine cross-validation

Use Cases

01Pre-commit validation to catch regressions and security flaws before they hit the repo

02Automating the initial pass of Pull Request reviews to ensure technical standards

03Ensuring architectural consistency and test coverage after major refactoring sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add guyathomas/workflows code-review-pipeline

For use in Claude.ai and ChatGPT

Download Skill