How does this differ from standard AI code reviews?

Unlike single-shot reviews, this skill uses a structured 7-phase workflow that gathers context, analyzes scope, and asks clarifying questions before performing the final analysis.

Does it require a Gemini API key?

Yes, this skill is designed to interface with the Gemini CLI to leverage large-context models like gemini-2.5-pro for comprehensive analysis.

Can it detect breaking changes?

Yes, Phase 2 of the workflow specifically analyzes the scope of changes to identify potential breaking API, database schema, or configuration modifications.

Are the code suggestions ready to implement?

The skill generates a detailed 'Fix Action Plan' in Phase 6, providing file-specific before/after code blocks and verification steps for each identified issue.

Can I customize the review criteria?

Absolutely. Phase 3 is a dedicated checkpoint where you can specify language standards, performance constraints, and specific architectural patterns for the AI to verify.

Comprehensive PR Reviewer

Name: Comprehensive PR Reviewer
Author: Expanly

byExpanly

0•

Herramientas de Colaboración

Automates a structured, multi-phase pull request review process using Gemini's large context for deep architectural and security analysis.

The Comprehensive PR Reviewer skill transforms standard code reviews into an iterative 7-phase workflow designed for high-stakes software development. By leveraging Gemini models, it moves beyond surface-level linting to perform deep analysis of business logic, architectural patterns, and security vulnerabilities. The process includes mandatory checkpoints where the AI gathers specific context on performance requirements and team standards, ensuring the final output provides actionable remediation plans, before-and-after code examples, and clear merge recommendations tailored to your project's unique constraints.

Características Principales

010 GitHub stars

02Automated change scope analysis across core logic, APIs, and databases

03Interactive checkpoints to align with team-specific architectural patterns

04Severity-based reporting ranging from critical blockers to positive highlights

05Actionable remediation plans with code examples and effort estimations

067-phase iterative workflow including context gathering and deep analysis

Casos de Uso

01Standardizing code review quality across distributed engineering teams

02Conducting deep security and logic audits on complex feature branches

03Identifying breaking changes and performance bottlenecks before merging

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add expanly/expanly-claude-code-agents comprehensive-pr-review

For use in Claude.ai and ChatGPT

Download Skill