How does this differ from standard AI code reviews?

Unlike basic reviews that check for syntax or style, this skill acts as a skeptical senior engineer, specifically looking for conceptual flaws, over-engineering, and hidden assumptions that LLMs often overlook.

What is the 'Multi-Claude' review pattern?

It involves running two separate Claude instances: one to write the code and a fresh one with this skill to review it. This ensures the reviewer is not biased by the implementation decisions made during the writing phase.

When is the best time to trigger this skill?

It should be used after generating significant amounts of code, before merging a pull request, or when dealing with security-sensitive and complex business logic.

What kind of output does the code review provide?

The skill generates a structured report covering critical issues, suggestions for improvement, identified assumptions, simplification opportunities, and clarifying questions.

Can this skill help reduce technical debt?

Yes, one of its primary functions is identifying over-engineering and 'clever' solutions where a simpler implementation would suffice, helping keep the codebase maintainable.

Skeptical Senior Code Review

Name: Skeptical Senior Code Review
Author: rahulsub

byrahulsub

•

コラボレーションツール

Performs rigorous code audits to identify hidden assumptions, over-engineering, and conceptual logic flaws from the perspective of a seasoned senior engineer.

This skill transforms Claude into a critical, high-level code auditor designed to catch the subtle conceptual errors that automated linters and standard AI reviews often miss. By adopting the mindset of a developer who has seen LLM-generated code fail in production, it scrutinizes pull requests for hidden assumptions, unnecessary complexity, and potential collateral damage. Whether used for self-review after a large code generation session or as a dedicated secondary 'reviewer' instance in a multi-agent workflow, it ensures your codebase remains lean, consistent, and resilient against edge cases.

主な機能

01Ensures architectural consistency and prevents collateral damage during refactors

02Provides a standardized review output including critical fixes and simplification opportunities

03Evaluates performance tradeoffs, maintenance implications, and potential failure modes

041 GitHub stars

05Detects over-engineering and suggests radical simplifications for complex solutions

06Identifies hidden assumptions and undocumented logic dependencies

ユースケース

01Auditing complex pull requests for logic flaws before merging into production

02Critiquing large blocks of AI-generated code to find subtle bugs and edge cases

03Implementing a 'Multi-Claude' review workflow with separate writer and reviewer instances

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rahulsub/code-practices-plugin code-review

For use in Claude.ai and ChatGPT

Download Skill