structure-audit

Audits codebase structure across 13 dimensions and generates a scored report with prioritized findings.

6|Updated Mar 27, 2026
One-click install
npx skills add https://github.com/greglas75/zuvo --skill structure-audit-greglas75
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: structure-audit
Source: https://github.com/greglas75/zuvo/tree/main/skills/structure-audit
Command: npx skills add https://github.com/greglas75/zuvo --skill structure-audit-greglas75

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Codebases accumulate structural debt over time: inconsistent naming, oversized files, dead code, duplication, and hidden hotspots. This Skill performs a quantitative structural health audit across 13 measurable dimensions (SA1-SA13), producing a scored report with letter grades, critical gates, and prioritized action items backed by tool evidence rather than subjective opinion. ## Core Features & Use Cases - 13-dimension structural analysis: Covers directory consistency, naming conventions, folder depth, colocation, barrel exports, separation of concerns, file size distribution, dead code, complexity, duplication, root organization, documentation, and churn-based hotspots. - Tool-driven evidence: Uses CodeSift as the primary analysis engine (detect_communities, analyze_complexity, find_dead_code, find_clones, analyze_hotspots) with CLI fallbacks like cloc, knip, dependency-cruiser, jscpd, and ESLint. - Scored report with gates: Assigns 0-3 scores per dimension, computes a normalized grade (A-D or FAIL), calculates a SQALE debt ratio, and enforces critical gates for layer violations, oversized files, and circular dependencies. - Use Case: Before a major refactor, run a full audit to identify god modules, circular dependencies, and high-churn hotspot files, then route findings to refactoring or architecture review workflows. ## Quick Start Ask the assistant to run a full structure audit on the current project, or scope it with a path or flags such as --naming, --dead-code, --hotspots, --quick, or --fix.

Frequently Asked Questions about structure-audit

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit codebase structure and organization?

Run the structure-audit skill with the full flag to evaluate 13 dimensions covering directory consistency, naming, file size, dead code, duplication, and hotspots. It produces a scored report with a letter grade and prioritized action items.

How to find dead code and unused exports in a TypeScript project?

Use the --dead-code flag to run CodeSift find_dead_code and find_unused_imports analysis, with knip as a CLI fallback. Results include unused export counts, unused file lists, and confidence levels per entry.

What tools detect code duplication and complexity?

This audit uses CodeSift find_clones and analyze_complexity as primary engines, falling back to jscpd for duplication and ESLint complexity rules or radon for Python. Duplication is flagged at 0.7 similarity or higher.

Does the structure audit work with Python or PHP projects?

Yes, stack detection supports Python via pyproject.toml or requirements.txt and PHP via composer.json. Python projects use radon and vulture fallbacks, while PHP projects use phpmd and phpcpd instead of JS-specific tools.

When should I use --quick mode versus a full audit?

Use --quick for fast checks that skip external tooling and git mining, covering only dimensions SA1-SA8 and SA11-SA12. Full mode adds complexity, duplication, and hotspot analysis but requires more tool execution time.

Why did my structure audit report get rejected?

Reports are rejected when findings lack file:line citations that resolve in the current tree, or when the retrospective entry is missing. Add a path/to/file.ext:LINE citation to every finding and re-run the append-runlog wrapper.