Design Token Audit Prompt

A prompt that analyzes your design token JSON and identifies issues like missing semantic layers, inconsistent naming, and accessibility gaps.

All Prompts

The Prompt

Analyze this design token JSON file and provide a detailed audit report.

Check for:
1. **Naming consistency** — Are tokens following a consistent naming convention? Flag any inconsistencies.
2. **Semantic layers** — Are there proper semantic tokens that reference primitives? Or are raw values used directly?
3. **Missing tokens** — Based on common design system patterns, what tokens are missing? (e.g., focus states, disabled states, error colors)
4. **Accessibility** — Do color pairings meet WCAG 2.1 AA contrast requirements? Flag any pairs that fail.
5. **Structure** — Is the token hierarchy logical? Are categories well-organized?

Format the output as:
- ❌ Critical issues (must fix)
- ⚠️ Warnings (should fix)
- 💡 Suggestions (nice to have)
- ✅ What's working well

Here's my token file:
[paste your tokens JSON here]

When to Use

  • Before a major design system release
  • When onboarding a new brand to a multi-brand system
  • During quarterly design system health checks
  • After migrating tokens between tools (Tokens Studio, Style Dictionary, etc.)

Expected Output

The AI will return a structured report with specific token names, line references, and suggested fixes. Typical audits catch 10-20 issues that would take hours to find manually.