CRAP score calculated with project baseline

Severity: Recommended

Check automatically:

1. Check for CRAP score tooling:

   # Node.js - look for complexity tools
   grep -E "complexity|crap|code-climate|plato|es6-plato" package.json 2>/dev/null
   
   # PHP - phpunit has CRAP built in
   grep -E "crap|coverage.*crap" phpunit.xml* 2>/dev/null
   
   # Python - look for radon or similar
   grep -E "radon|mccabe|complexity" requirements*.txt pyproject.toml 2>/dev/null
   

2. Check for baseline/threshold configuration:

   # Look for documented thresholds
   grep -riE "crap.*threshold|crap.*baseline|complexity.*threshold|max.*complexity" \
     README.md CLAUDE.md docs/ .eslintrc* package.json 2>/dev/null
   

3. Run CRAP/complexity analysis (if tooling exists):

   # Node.js with eslint complexity rule
   pnpm eslint --rule 'complexity: ["error", 10]' src/
   
   # PHP
   ./vendor/bin/phpunit --coverage-text | grep -A 20 "CRAP"
   
   # Python
   radon cc src/ -a -s
   

Cross-reference with:

• TEST-001 - CRAP score often calculated alongside coverage
• TODO: Section 28 (Code Architecture) - update reference once developed
• TODO: Section 40 (Technical Debt) - update reference once developed

Pass criteria:

• CRAP score or complexity tooling configured
• Baseline/threshold established for the project
• Current scores within acceptable range

Fail criteria:

• No complexity measurement tooling
• No baseline defined (can't track if it's getting worse)
• High CRAP scores ignored

If no tooling found, ask user: "No CRAP score or complexity measurement tooling found. Do you want to track code complexity? Common options: ESLint complexity rule, PHPUnit CRAP, radon (Python)."

Evidence to capture:

• Tooling detected (or none)
• Baseline/threshold value (if configured)
• Current CRAP/complexity scores for key components
• Components exceeding threshold (if any)

---