How does the Code Smell Detector trigger?

It activates automatically whenever you use Write or Edit tools in Claude Code, or specifically when you request a review using keywords like 'review' or 'refactor'.

Does it support automatic code fixing?

Yes, for many common smells, you can use the 'smell fix' command to have Claude automatically apply the recommended refactoring patterns.

Can I customize the detection thresholds?

Yes, you can define specific thresholds for maximum function lines, parameter counts, and indentation levels to match your team's coding standards.

How is the code health grade calculated?

The skill assigns a severity score (Critical: 10pts, Warning: 5pts, Info: 2pts). Total points determine a grade from A (Clean) to F (Severe), helping you assess technical debt at a glance.

Which specific code smells can this skill detect?

The skill detects 22 patterns including God Objects, Long Methods, Duplicate Code, Feature Envy, Data Clumps, Primitive Obsession, and Magic Numbers.

Code Smell Detector

Name: Code Smell Detector
Author: monicajeon28

bymonicajeon28

開発者ツール

Detects 22 types of code smells and provides immediate refactoring suggestions to prevent technical debt.

概要

The Code Smell Detector is a specialized Claude Code skill designed to maintain high code quality by identifying problematic patterns like God Objects, Long Methods, and Feature Envy. It acts as an automated static analysis tool that triggers during file edits or pull requests, calculating a severity score and providing actionable refactoring advice. By catching issues early—from critical architectural flaws to minor magic numbers—it helps developers reduce technical debt and ensure long-term maintainability through a standardized grading system (A-F).

主な機能

Real-time feedback loops triggered during code creation, editing, and review
Automatic detection of 22 distinct code smell patterns including God Object and Data Clumps
Weighted severity scoring system to help prioritize high-impact refactoring
0 GitHub stars
Configurable complexity thresholds for file length, nesting depth, and parameter counts
Detailed 'Before vs. After' refactoring examples for every detected issue

ユースケース

Automated code reviews to identify architectural flaws before merging pull requests
Proactive refactoring during development to prevent the accumulation of technical debt
Legacy code auditing to generate a prioritized 'hit list' of areas requiring cleanup

概要

主な機能

Real-time feedback loops triggered during code creation, editing, and review
Automatic detection of 22 distinct code smell patterns including God Object and Data Clumps
Weighted severity scoring system to help prioritize high-impact refactoring
0 GitHub stars
Configurable complexity thresholds for file length, nesting depth, and parameter counts
Detailed 'Before vs. After' refactoring examples for every detected issue

ユースケース

Automated code reviews to identify architectural flaws before merging pull requests
Proactive refactoring during development to prevent the accumulation of technical debt
Legacy code auditing to generate a prioritized 'hit list' of areas requiring cleanup