refactor-audit

name	refactor-audit
description	Audit codebase for refactoring candidates. Identifies complexity issues, large files/functions, code duplication, and maintainability problems. Outputs report to docs/reports/.
disable-model-invocation	true
argument-hint	[scope] [--quick]
automation	manual
metadata	{"version":"1.0","created":"2026-03-13T00:00:00.000Z","author":"Ability.ai","changelog":["1.0: Initial version — audits the codebase for refactoring candidates (complexity, large files/functions, duplication) and writes a report to docs/reports/."]}

Refactor Audit

ℹ️ First, set expectations: before anything else, print one short line with this skill's version and its most recent change — the top entry of metadata.changelog above — e.g. refactor-audit vX.Y — recent: <summary>. Then proceed.

Analyze code to identify refactoring candidates with complexity-based thresholds.

State Dependencies

Source	Location	Read	Write	Description
Source Code	src/	✅		Files to analyze
Audit Report	docs/reports/refactor-audit-{date}.md		✅	Generated report

Usage

/refactor-audit                    # Full src/ scan
/refactor-audit backend            # Only src/backend/
/refactor-audit frontend           # Only src/frontend/
/refactor-audit src/backend/app.py # Single file
/refactor-audit --quick            # Top 10 issues only

Arguments

• $ARGUMENTS - Scope (backend/frontend/path) and flags

Procedure

Phase 1: Parse Arguments

Determine scope from $ARGUMENTS:

• Empty or "all" -> src/
• "backend" -> src/backend/
• "frontend" -> src/frontend/
• Path (contains / or file extension) -> use as-is
• "--quick" flag -> limit to top 10 issues

Phase 2: Run Analysis

For Python files:

find [scope] -name "*.py" -exec wc -l {} \; | awk '$1 > 500 {print}'

For JavaScript/TypeScript files:

find [scope] -name "*.js" -o -name "*.ts" -o -name "*.vue" -o -name "*.tsx" | xargs wc -l | awk '$1 > 400 {print}'

Phase 3: Categorize Findings

P0 Critical - Must fix:

• Files >1000 lines
• Functions >200 lines
• Cyclomatic complexity >30

P1 High - Strongly recommended:

• Files 800-1000 lines
• Functions 100-200 lines
• Significant duplication (>10 similar lines)

P2 Medium - Recommended:

• Files 500-800 lines
• Functions 50-100 lines
• Parameters >7
• Nesting depth >4

P3 Low - Nice to have:

• Files 300-500 lines
• Functions 30-50 lines
• Minor duplication (6-10 lines)

Phase 4: Generate Report

Save report at docs/reports/refactor-audit-YYYY-MM-DD.md with summary table, critical issues, hotspots, and recommendations.

Phase 5: Report Location

Confirm report saved and provide summary with top 3 hotspots.

Completion Checklist

• Scope correctly parsed from arguments
• Analysis run (complexity, file sizes, duplication)
• Findings categorized by severity (P0-P3)
• Hotspots identified (files with multiple issues)
• Report saved to docs/reports/
• Summary output provided