Which expert personas can I use for the critique?

You can choose from grug, carmack, ousterhout, fowler, beck, jobs, and torvalds, each offering a unique lens on code quality.

How does the adversarial review help my development?

It acts as a ruthless second pair of eyes that challenges your assumptions, helping you catch over-engineering and design flaws before they reach production.

What kind of output does the skill provide?

It provides a structured critique detailing what the persona hates, what they demand you change, and the specific conditions under which they would approve the work.

Can I use this for high-level architectural designs?

Yes, the skill is highly effective for reviewing both raw source code and high-level design documents, RFCs, or project plans.

Is the feedback purely negative?

While adversarial in nature, the feedback is designed to be constructive by offering concrete suggestions and fixes to improve the robustness and simplicity of your code.

Expert Code Critique

Name: Expert Code Critique
Author: phrazzld

byphrazzld

•

协作工具

Provides adversarial code reviews by channeling industry legends to identify architectural flaws, complexity, and design smells.

The Critique skill transforms Claude into an adversarial reviewer, allowing developers to subject their code, designs, or plans to the scrutiny of specific industry personas like John Carmack, Martin Fowler, or Linus Torvalds. By applying specialized lenses—such as grug's focus on simplicity or Kent Beck's emphasis on testability—the skill ruthlessly identifies over-engineering, technical debt, and 'complexity demons,' providing concrete recommendations to align your work with world-class engineering standards.

主要功能

01Structured output highlighting specific demands and approval conditions

02Pragmatic focus on shippability and scope creep prevention

03Persona-based feedback from legends like Carmack, Fowler, and Torvalds

045 GitHub stars

05Identification of code smells, duplication, and feature envy

06Deep analysis of architectural complexity and over-abstraction

使用场景

01Auditing test suites for over-mocking and poor TDD patterns

02Reviewing new feature plans for scope creep before beginning implementation

03Refactoring complex modules to reduce over-abstraction and 'big-brain' patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add phrazzld/claude-config critique

For use in Claude.ai and ChatGPT

Download Skill