claude-design-auditor-skill

Design Auditor is a Claude skill that audits UI/design inputs against 18 categories (including accessibility/WCAG and ethics/dark patterns). It can analyze Figma files via Figma MCP, as well as code (HTML/CSS/React/Vue), screenshots, and written descriptions, producing scored reports, prioritized issues, and developer handoff/export artifacts (e.g., Markdown, before/after diffs). It supports English and Korean.

Evaluated Mar 30, 2026 (22d ago)
Repo ↗ Ai Ml ai-ml claude-skill design-audit accessibility wcag figma design-tokens frontend ethics dark-patterns
⚙ Agent Friendliness
46
/ 100
Can an agent use this?
🔒 Security
39
/ 100
Is it safe for agents?
⚡ Reliability
44
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
55
Documentation
75
Error Messages
0
Auth Simplicity
60
Rate Limits
10

🔒 Security

TLS Enforcement
60
Auth Strength
35
Scope Granularity
30
Dep. Hygiene
35
Secret Handling
35

The provided content does not specify transport (TLS) enforcement, auth mechanism, scope model, or secret-handling practices. Given it integrates with Figma via MCP, authorization and least-privilege should be verified, but the README does not document these details. The skill is designed to analyze code and design assets; ensure that any included artifacts are handled according to the hosting platform’s privacy/security policies. No dependency/SBOM or CVE hygiene information is provided.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
70
Error Recovery
50
AF Security Reliability

Best When

You want a developer-friendly, rule-based second opinion on design quality (including WCAG-aligned accessibility and ethics/dark-pattern heuristics) for web UIs and design-system components, especially when you can provide Figma links or code/screenshots.

Avoid When

You need definitive legal/regulatory compliance assurance, or you cannot provide enough artifacts (e.g., only vague descriptions) for the rule checks to be reliable.

Use Cases

  • Audit UI designs for typography, color/contrast, spacing/layout, and visual hierarchy
  • Run accessibility-focused checks (WCAG-related, including touch targets, focus states, labels, reading order)
  • Assess ethics/dark-pattern risks (e.g., deceptive UX, coercive flows, consent/privacy issues)
  • Review code for design rule compliance and generate before/after diffs for fixes
  • Validate Figma designs and (optionally) apply fixes in-place via Figma MCP
  • Create developer handoff reports (CSS spec tables, accessibility checklists, critical fixes)
  • Export audit results as Markdown for Notion/GitHub/Linear/Jira

Not For

  • Replacing professional accessibility audits or legal/consumer-rights review for compliance claims
  • Doing exhaustive UX research or usability testing with real users
  • Designing brand systems requiring strong brand/visual identity governance beyond rule-based checks
  • Guaranteeing that an interface is free from dark patterns or fully compliant with all jurisdictional requirements

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: Claude Skill installation via Claude.ai (no explicit auth method described in README) Figma MCP workflow for accessing Figma resources (credentials/authorization not described in README)
OAuth: No Scopes: No

No API authentication details are provided in the provided README for the skill itself. Figma MCP is mentioned as an integration path, but the specific auth mechanism (OAuth scopes, tokens, domain restrictions) is not documented in the supplied content.

Pricing

Free tier: No
Requires CC: No

Pricing is not described in the provided README content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Figma MCP step/workflow is described at a high level; tool/operation granularity and failure modes (e.g., partial page selection, variable access errors) are not detailed in the supplied README.
  • Because the skill performs edits in Figma 'via MCP' (claimed), agents should be cautious about unintended edits and confirm scope/permissions before applying fixes.
  • The repository content provided does not include explicit machine-readable error codes, so agent retries/error handling may be less deterministic than with a typical API.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for claude-design-auditor-skill.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered