Code Complexity Analyzer

Every codebase has hidden landmines — functions so tangled that modifying them is a coin flip between fix and regression. This skill finds them systematically.

The Code Complexity Analyzer measures cyclomatic complexity (how many paths), cognitive complexity (how hard to understand), maintainability index (overall health score), and cross-references with git churn data to surface the files that are both complex AND change frequently — your highest-risk targets.

Works with any language. Includes tool-based analysis for JavaScript, TypeScript, Python, Go, and Java, plus a universal manual analysis method that works by reading source files directly — no dependencies required.

What you get:

• Complete metrics reference with severity thresholds and action triggers
• Hotspot report format that ranks every function by risk score
• Churn x Complexity matrix identifying your most dangerous files
• Five refactoring patterns with before/after validation criteria
• CI integration examples (GitHub Actions, package.json scripts)
• Trend tracking to measure improvement over time

The risk scoring formula (CC x Cognitive x log(Lines)) surfaces the functions that combine multiple complexity signals. A function with CC of 30, cognitive complexity of 40, and 180 lines scores dramatically higher than one with CC of 15 and 30 lines — because the compound effect of multiple complexity dimensions is what makes code truly unmaintainable.

Designed for engineering leads making refactoring decisions, developers preparing tech debt proposals, and teams establishing complexity budgets. Every recommendation includes specific action — not just "this is complex" but "extract validation into a separate function" or "replace this switch with a strategy pattern."

No external services. No API keys. Works on any codebase your agent can read.