code-review-excellence

Structure pull request reviews with severity-prefixed, actionable comments.

3|Updated Feb 9, 2026
One-click install
npx skills add https://github.com/marcus-friction/agents --skill code-review-excellence-marcus-friction
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-review-excellence
Source: https://github.com/marcus-friction/agents/tree/main/.agent/skills/code-review-excellence
Command: npx skills add https://github.com/marcus-friction/agents --skill code-review-excellence-marcus-friction

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Improve the quality and consistency of code reviews by providing meta-level guidance and a structured feedback approach.

Core Features & Use Cases

  • One-concern-per-comment guideline to maintain clarity and focus.
  • Severity prefixes (🔴 Must, 🟡 Should, 🟢 Nit, ❓ Question) to communicate importance.
  • Stepwise review order (Understand, Correctness, Standards, Security, Performance, Architecture, Accessibility, Tests, Report).
  • Consistent feedback templates for actionable improvement across architectures, security, and style.

Quick Start

Apply this guide to a pending PR by reading the description, tracing the code flow, and leaving one clear, actionable comment per issue with a suggested fix.

Frequently Asked Questions about code-review-excellence

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I structure pull request feedback for better code reviews?▼

Structure pull request feedback by leaving one clear, actionable comment per issue with a suggested fix. Use severity prefixes like Must, Should, Nit, and Question to communicate importance clearly.

What is the best order to review code across architecture and security dimensions?▼

The best order to review code is stepwise: Understand, Correctness, Standards, Security, Performance, Architecture, Accessibility, and Tests. This sequence ensures comprehensive coverage before generating the final report.

Can I apply this code review guidance to any programming language or framework?▼

Yes, you can apply this code review guidance to any language or framework. It provides meta-level guidance covering architecture, correctness, security, performance, and accessibility dimensions universally.

How do I communicate severity levels during a code review?▼

Communicate severity levels during a code review using specific prefixes: Must for critical issues, Should for important improvements, Nit for minor style points, and Question for inquiries.

Why maintain a one-concern-per-comment guideline in code review feedback?▼

Maintain a one-concern-per-comment guideline to preserve clarity and focus. Isolated feedback ensures developers can address individual issues systematically without getting lost in mixed suggestions.