How does Deep Review differ from a standard linter?

Unlike basic linters, Deep Review uses specialized AI agents to analyze high-level concepts like architecture, logic errors, concurrency issues, and platform-specific best practices that static tools often miss.

How are 'Pre-existing' issues handled?

The skill distinguishes between [NEW] issues introduced in your current work and [PRE-EXISTING] issues already in the file, allowing you to prioritize fixes while maintaining overall module health.

Does it support my specific programming language?

It features intelligent auto-detection for over 30 platforms including TypeScript, Python, Go, Rust, iOS, Android, and infrastructure tools like Terraform and Docker.

Can I run a review on just my current changes?

Yes, you can use the --changes flag to analyze only uncommitted work or the --pr flag to review everything modified in your current branch relative to the base.

Deep Review

Name: Deep Review
Author: Iron-Ham

byIron-Ham

•

보안 및 테스팅

Performs comprehensive, multi-agent code reviews across architecture, security, performance, and platform-specific best practices.

Deep Review is an advanced auditing tool for Claude Code that deploys a team of specialized AI agents to conduct thorough assessments of your codebase. It analyzes everything from architectural patterns and memory leaks to accessibility (a11y) and PII exposure, automatically detecting the relevant tech stack—ranging from Next.js and iOS to Rust and Terraform—to provide context-aware feedback. Whether you are preparing a pull request or tackling technical debt, this skill distinguishes between newly introduced bugs and inherited issues, ensuring a high standard of code health and maintainability across your entire project.

주요 기능

01Deep architectural analysis identifying dependency cycles, hotspots, and scalability concerns.

02Flexible scope targeting using Git flags for pull requests, specific branches, or uncommitted changes.

03Automatic platform detection for 30+ frameworks and languages like React, Swift, Go, and Kubernetes.

04Multi-agent review system covering 15+ specialized aspects including security, performance, and concurrency.

058 GitHub stars

06Intelligent issue classification distinguishing between [NEW] changes and [PRE-EXISTING] technical debt.

사용 사례

01Conducting a final comprehensive audit before merging a critical pull request to production.

02Running automated security and PII leak scans to ensure compliance with data protection standards.

03Identifying and categorizing technical debt in legacy modules during a major refactoring phase.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add iron-ham/claude-deep-review deep-review

For use in Claude.ai and ChatGPT

Download Skill