design-system-audit

Audits a design system codebase across dozens of quality dimensions and remediates findings.

5|9|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/ajenchen/design-system --skill design-system-audit-ajenchen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: design-system-audit
Source: https://github.com/ajenchen/design-system/tree/main/.agents/skills/design-system-audit
Command: npx skills add https://github.com/ajenchen/design-system --skill design-system-audit-ajenchen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Design systems accumulate drift between specs, code, tokens, and stories over time, and manual review misses recurring bug classes. This Skill runs a systematic, dimension-by-dimension audit of a design system repository and remediates actionable findings under a shared authority classifier. ## Core Features & Use Cases - Multi-dimension audit registry: Runs grouped audit dimensions covering cva defaultVariants drift, SSOT dead links, token consumption discipline, spec hygiene, shadcn passthrough completeness, a11y coverage, story layer quality, naming consistency, and layout family compliance. - Subagent prompt library: Ships per-dimension prompts in references/audit-prompts.md with Absolute vs Consistency classification, false-positive memory, and canonical source citations so findings are evidence-based. - Authority-classified remediation: Applies P0/P1 fixes autonomously and only stops for genuine product/UI/UX tradeoffs, using checkpoint templates in references/checkpoints.md. - Use Case: Invoke when asked to audit, re-audit, or verify design system health before a release or after a large token refactor; the Skill dispatches dimension checks, consolidates findings by file and line, and fixes engineering-class issues directly. ## Quick Start Ask the assistant to run a full design system audit on the repository and report prioritized findings with file and line references.

Frequently Asked Questions about design-system-audit

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

FAQPage Schema
How do I run a design system audit on my component library?

Invoke the audit when you want to check quality, verify design system health, or prepare for a release. It runs the full dimension registry covering spec hygiene, code correctness, tokens, naming, and stories, then reports prioritized findings with file and line references.

What kinds of bugs does a design system audit catch?

It catches cva defaultVariants three-way drift between code, spec, and anatomy stories, dead SSOT links, hardcoded hex or pixel values where tokens exist, missing forwardRef or displayName, missing aria-labels, and placeholder text in stories.

Does the audit fix issues automatically or only report them?

It remediates P0 and P1 engineering findings autonomously under a standing authorization model. It only pauses for genuine product, UI, or UX tradeoffs that require a human decision, presenting those as batched checkpoints.

How does the audit avoid false positives on consistency checks?

Each dimension is classified as Absolute or Consistency, and subagents must grep the component spec for documented rationale and check scope defaults before flagging a violation. A curated false-positive memory list in the references prevents repeat mistakes.

When should I not use this audit skill?

Do not use it for consumer-layer application UI review, which belongs to a separate product UI audit scope. It is designed for the design system package internals such as specs, tokens, primitives, and stories, not end-user app code.