Does this skill provide praise for good code?

No, this specific skill focuses on identifying and classifying findings; it purposefully filters out praise and PR metadata to keep the review concise and actionable.

What are the five severity levels used for classification?

Findings are classified as CRITICAL (breaks/crashes), IMPORTANT (bugs/error handling), DEBT (patterns/conventions), SUGGESTED (measurable improvements), or QUESTION (unclear intent).

How does this skill handle subjective 'nitpicks' in code reviews?

The skill is configured to reject vague suggestions, naming nitpicks, and style preferences without an enforced standard to focus on high-impact feedback.

What counts as a 'Suggested Improvement' (🎨)?

A suggestion must provide measurable value, such as reducing cyclomatic complexity by 3+, eliminating a vulnerability class, or significantly improving performance.

Review Findings Classifier

Name: Review Findings Classifier
Author: bitwarden

bybitwarden

•

Herramientas de Colaboración

Categorizes code review findings into standardized severity levels using objective metrics and hybrid emoji formatting.

The Classifying Review Findings skill provides a rigorous framework for organizing PR feedback into five distinct severity categories: Critical, Important, Debt, Suggested, and Question. It helps maintain code quality by enforcing a strict verification process before a finding is logged, ensuring all suggestions are backed by measurable evidence such as cyclomatic complexity reduction or security gains. By filtering out subjective nitpicks and vague praise, this skill streamlines communication between developers and ensures that the most impactful issues are addressed first during the review process.

Características Principales

01Automatic filtering of subjective nitpicks, praise, and vague suggestions

02Hybrid emoji and text formatting for immediate visual prioritization in PRs

031 GitHub stars

04Metric-driven criteria for suggested improvements based on complexity and security

05Strict verification logic to prevent unconfirmed or false-positive findings

06Standardized severity mapping from Critical (❌) to Question (❓)

Casos de Uso

01Reducing code review noise by enforcing measurable improvement standards

02Automating high-signal security and performance audits in CI/CD workflows

03Standardizing PR feedback across distributed engineering teams

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bitwarden/ai-plugins classifying-review-findings

For use in Claude.ai and ChatGPT

Download Skill