---

name: quality-check description: This skill should be used when the user asks to "check code quality", "score this code", "find quality hotspots", "prioritize refactoring", "analyze technical debt", "run CodeDNA", or assess maintainability, complexity, modularity, tests, documentation, and best practices. version: 0.1.0 source: fork checksum: 0fee4661b24036455cab84e6e48678ac7a5cd8a794f10d4550d56452933d8cc6 updated_at: 2026-06-08T10:50:00+08:00 layer: quality

Codex CLI: Invoke when the description matches, or manually with $quality-check. No hooks or background auto-run.

Code Quality Check Skill

When to Use This Skill

Automatically invoke this Skill when:

• User asks to "check code quality", "review this code"
• User mentions refactoring or code improvements
• User wants quality scores or metrics
• Before major refactoring efforts
• During code review processes
• Keywords: "quality", "refactor", "review", "improve code", "code smell"

What This Skill Does

CodeDNA Quality Analyzer provides:

1. 6-Dimension Scoring - Comprehensive quality assessment
2. Issue Identification - Specific problems with severity levels
3. ROI-Optimized Suggestions - High-value refactoring priorities
4. PageRank Analysis - Quality hotspot identification

The 6 Quality Dimensions

1. Complexity - Cyclomatic complexity, nesting depth
2. Maintainability - Code readability, documentation quality
3. Modularity - Coupling, cohesion, dependency structure
4. Test Coverage - Test quality and coverage metrics
5. Documentation - Comment ratio, API docs completeness
6. Best Practices - Style guide adherence, pattern usage

Instructions

When this Skill is invoked:

Step 1: Determine Analysis Mode

Ask yourself:

• Is it a single file? → Use test_quality_simple.py
• Is it a project directory? → Use project_quality_analyzer.py --mode overview
• Need impact analysis? → Use project_quality_analyzer.py --mode impact-analysis

Step 2: Resolve Runtime and Execute the Analysis

IMPORTANT: Execute one of these analyzer commands. Prefer python3, fall back to python, and allow PROJECTMIND_HOME to override the default local tool path.

Single File Analysis:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
if [ -z "$PYTHON_BIN" ]; then
  echo "CodeDNA error: no python3 or python interpreter found" >&2
  exit 1
fi
"$PYTHON_BIN" "$PROJECTMIND_HOME/test_quality_simple.py" "[file_path]"

Project Overview:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
if [ -z "$PYTHON_BIN" ]; then
  echo "CodeDNA error: no python3 or python interpreter found" >&2
  exit 1
fi
"$PYTHON_BIN" "$PROJECTMIND_HOME/project_quality_analyzer.py" --project "[directory_path]" --mode overview

Impact Analysis:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
if [ -z "$PYTHON_BIN" ]; then
  echo "CodeDNA error: no python3 or python interpreter found" >&2
  exit 1
fi
"$PYTHON_BIN" "$PROJECTMIND_HOME/project_quality_analyzer.py" --file "[file_path]" --mode impact-analysis

If $PYTHON_BIN is empty, report that no Python interpreter was found and fall back to direct local inspection with rg, rg --files, targeted reads, and the repo's own test/lint commands.

Step 3: Present Results

Format the output as:

## 🔬 Code Quality Analysis

**Target**: [File/Project Path]
**Overall Score**: XX/100 [🟢/🟡/🔴]

### 6-Dimension Breakdown

| Dimension | Score | Status | Key Insights |
|-----------|-------|--------|--------------|
| Complexity | XX/100 | 🟢/🟡/🔴 | [Main issue] |
| Maintainability | XX/100 | 🟢/🟡/🔴 | [Main issue] |
| Modularity | XX/100 | 🟢/🟡/🔴 | [Main issue] |
| Test Coverage | XX/100 | 🟢/🟡/🔴 | [Main issue] |
| Documentation | XX/100 | 🟢/🟡/🔴 | [Main issue] |
| Best Practices | XX/100 | 🟢/🟡/🔴 | [Main issue] |

### 🔴 Critical Issues (Fix Immediately)
1. **[Issue Name]** - [File:Line]
   - Impact: [High/Medium/Low]
   - Explanation: [What's wrong]
   - Fix: [Specific solution]

### 🟡 Important Issues (Fix This Sprint)
1. **[Issue Name]** - [File:Line]
   - Impact: [explanation]
   - Suggested approach: [how to fix]

### 💡 Refactoring Priorities (ROI-Optimized)

Based on PageRank and quality analysis:

1. **[Module/File Name]** (ROI: High)
   - Current Score: XX/100
   - Effort: [X hours]
   - Benefit: [Specific improvements]
   - Priority: P0/P1/P2

2. **[Module/File Name]** (ROI: Medium)
   - [Details...]

### 📊 Quality Hotspots

Files that would benefit most from refactoring:
- [File 1]: Score XX, high dependency count
- [File 2]: Score XX, complex and frequently changed
- [File 3]: Score XX, critical business logic

### 🎯 Recommended Actions

**Immediate (This Week)**:
- [ ] [Specific action with file:line]
- [ ] [Specific action with file:line]

**Short-term (This Sprint)**:
- [ ] [Refactoring task]
- [ ] [Testing improvement]

**Long-term (Next Quarter)**:
- [ ] [Architecture improvement]
- [ ] [Technical debt reduction]

Step 4: Provide Context

Explain:

• Why these scores matter
• Business impact of the issues
• Risk of not fixing critical problems
• Expected improvement from suggested changes

Examples

Example 1: Single File Review

User: "Check the quality of src/auth/login.ts"

Execute:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
"$PYTHON_BIN" "$PROJECTMIND_HOME/test_quality_simple.py" src/auth/login.ts

You present: 6-dimension scores, specific issues, and refactoring suggestions.

Example 2: Project Overview

User: "How's the code quality of my payment module?"

Execute:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
"$PYTHON_BIN" "$PROJECTMIND_HOME/project_quality_analyzer.py" --project src/payment --mode overview

You present: Project-wide quality assessment, hotspots, and priority fixes.

Example 3: Impact Analysis

User: "I want to refactor the database layer, what's the impact?"

Execute:

PROJECTMIND_HOME="${PROJECTMIND_HOME:-/Users/WangQiao/claude-enhanced-quality}"
PYTHON_BIN="${PYTHON_BIN:-$(command -v python3 || command -v python || true)}"
"$PYTHON_BIN" "$PROJECTMIND_HOME/project_quality_analyzer.py" --file src/core/database.ts --mode impact-analysis

You present: Dependency analysis, affected files, risk assessment, and refactoring plan.

Quality Thresholds

Scoring system:

• 🟢 80-100: Good quality, minor improvements only
• 🟡 60-79: Acceptable, needs improvement
• 🔴 <60: Poor quality, refactoring required

ROI Calculation

Refactoring priority is based on:

• Quality Score (lower = higher priority)
• PageRank (higher = more important)
• Dependency Count (higher = more impact)
• Change Frequency (higher = more value)

Integration with ProjectMind

For project-level analysis, the system uses:

• Knowledge Graph - Deep project understanding
• Dependency Mapping - Complete relationship analysis
• Historical Data - Evolution patterns and trends

Important Notes

• Always execute the Python command, don't guess scores
• Explain the "why" behind each issue
• Prioritize by ROI, not just severity
• Provide specific fixes, not generic advice
• Consider business context in recommendations

Common Quality Issues

Complexity Issues

• High cyclomatic complexity (>10)
• Deep nesting (>4 levels)
• Long functions (>50 lines)
• God classes/objects

Maintainability Issues

• Poor naming conventions
• Missing documentation
• Magic numbers/strings
• Code duplication

Modularity Issues

• High coupling
• Low cohesion
• Circular dependencies
• Tight integration

Testing Issues

• Low coverage (<80%)
• Missing edge cases
• Flaky tests
• No integration tests

Prerequisites

• Python environment with CodeDNA installed
• Access to project files
• ProjectMind/CodeDNA tools installed at /Users/WangQiao/claude-enhanced-quality or another path supplied through PROJECTMIND_HOME

Performance

• Single File: <5 seconds
• Project Overview: 10-30 seconds
• Impact Analysis: 15-45 seconds