home / skills / thebushidocollective / han / figma-extract-tokens

This skill extracts design tokens from Figma into multiple formats to empower scalable design systems.

npx playbooks add skill thebushidocollective/han --skill figma-extract-tokens

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
5.5 KB
---
name: figma-extract-tokens
description: Extract design tokens and variables from a Figma file to create or update a design system
---

# Extract Design Tokens from Figma

## Name

figma:figma-extract-tokens - Extract design tokens and variables from Figma files

## Synopsis

Extract design tokens (colors, spacing, typography, etc.) from Figma files and generate token files in various formats (CSS Custom Properties, TypeScript, JSON, Style Dictionary, Tailwind config).

## Description

You are tasked with extracting design tokens and variables from a Figma file using the Figma MCP server. This command analyzes Figma variables and variable collections to generate structured token files for your design system.

## Implementation

Connects to the Figma Desktop MCP server to access Figma variables API. Extracts all variable types (colors, numbers, strings, booleans) and organizes them into primitive and semantic token hierarchies. Supports multiple output formats and theming variations.

## Your Task

1. **Access the Figma File**:
   - Use Figma MCP tools to access the current file
   - If no file is open, ask the user to open one in Figma or provide a URL

2. **Extract Variables**:
   - Get all color variables and their values
   - Get spacing/sizing variables
   - Get typography variables (font families, sizes, weights, line heights)
   - Get any other variables (border radius, shadows, etc.)
   - Organize by variable collections if they exist

3. **Analyze Token Structure**:
   - Identify semantic tokens (primary, secondary, etc.)
   - Note primitive tokens (specific values)
   - Detect theming variations (light/dark modes)
   - Check for token hierarchies and references

4. **Generate Token Files**:
   - Ask the user what format they prefer:
     - CSS Custom Properties
     - JavaScript/TypeScript objects
     - JSON
     - Sass/SCSS variables
     - Tailwind config
     - Style Dictionary format
   - Generate properly structured token files
   - Include comments with Figma references

5. **Provide Documentation**:
   - Create a mapping of token names to Figma variables
   - Note any missing tokens that should be added
   - Suggest token naming conventions if inconsistent
   - Document how to use tokens in code

## Output Formats

### CSS Custom Properties

```css
/**
 * Design Tokens - Extracted from Figma
 * File: [Figma file name]
 * Generated: [date]
 */

:root {
  /* Colors - Primitives */
  --color-blue-500: #3b82f6;
  --color-gray-900: #111827;

  /* Colors - Semantic */
  --color-primary: var(--color-blue-500);
  --color-text: var(--color-gray-900);

  /* Spacing */
  --spacing-xs: 0.25rem;
  --spacing-sm: 0.5rem;
  --spacing-md: 1rem;

  /* Typography */
  --font-family-base: 'Inter', sans-serif;
  --font-size-body: 1rem;
  --line-height-body: 1.5;
}
```

### TypeScript/JavaScript

```typescript
/**
 * Design Tokens - Extracted from Figma
 * File: [Figma file name]
 * Generated: [date]
 */

export const tokens = {
  colors: {
    primitives: {
      blue500: '#3b82f6',
      gray900: '#111827',
    },
    semantic: {
      primary: '#3b82f6',
      text: '#111827',
    },
  },
  spacing: {
    xs: '0.25rem',
    sm: '0.5rem',
    md: '1rem',
  },
  typography: {
    fontFamily: {
      base: "'Inter', sans-serif",
    },
    fontSize: {
      body: '1rem',
    },
    lineHeight: {
      body: 1.5,
    },
  },
} as const;

export type Tokens = typeof tokens;
```

### Style Dictionary Format

```json
{
  "color": {
    "primitive": {
      "blue": {
        "500": { "value": "#3b82f6" }
      },
      "gray": {
        "900": { "value": "#111827" }
      }
    },
    "semantic": {
      "primary": { "value": "{color.primitive.blue.500}" },
      "text": { "value": "{color.primitive.gray.900}" }
    }
  },
  "spacing": {
    "xs": { "value": "0.25rem" },
    "sm": { "value": "0.5rem" },
    "md": { "value": "1rem" }
  }
}
```

## Best Practices

1. **Token Naming**:
   - Use semantic names for application tokens (primary, secondary)
   - Use descriptive names for primitive tokens (blue-500, spacing-md)
   - Be consistent with naming conventions
   - Use namespacing for organization

2. **Token Organization**:
   - Group related tokens (colors, spacing, typography)
   - Separate primitives from semantics
   - Support theming with token references
   - Document token purposes

3. **Token Values**:
   - Preserve units from Figma (px, rem, em)
   - Convert to appropriate format for target platform
   - Include fallbacks for token references
   - Validate values are valid CSS/code

4. **Maintenance**:
   - Include Figma file reference for traceability
   - Add generation timestamp
   - Note any manual adjustments needed
   - Document how to re-sync from Figma

## Additional Deliverables

After extracting tokens, also provide:

1. **Token Documentation**:
   - Markdown file documenting each token category
   - Usage examples for designers and developers
   - Migration guide if updating existing tokens

2. **Integration Guide**:
   - How to import tokens in the project
   - How to use tokens in components
   - How to extend or override tokens

3. **Validation**:
   - Check for duplicate tokens
   - Note any missing semantic tokens
   - Identify unused variables in Figma
   - Suggest token consolidation opportunities

## Notes

- If variable collections exist, extract them separately
- Note any Figma variables without values
- Identify color mode variations (light/dark)
- Suggest token additions for common use cases not covered
- Provide examples of using tokens in actual components

Overview

This skill extracts design tokens and variables from an open Figma file and generates ready-to-use token artifacts for design systems. It reads Figma variables and collections, organizes primitives and semantic tokens, and emits files in formats like CSS custom properties, TypeScript, JSON, Style Dictionary, Tailwind, or Sass. The output includes mapping back to Figma, generation metadata, and suggestions to improve token consistency.

How this skill works

The skill connects to the Figma MCP (desktop) variables API to read all variable types (colors, numbers, strings, booleans) and groups by collection. It analyzes references to detect primitive vs semantic tokens, theming variations (light/dark), and hierarchical relationships. After analysis it generates structured token files in the requested format and produces documentation: a name-to-Figma mapping, missing-token notes, and usage examples with Figma references and timestamps.

When to use it

  • Onboard an existing Figma file into a design system
  • Export tokens for front-end implementation (CSS, JS/TS, Tailwind, Style Dictionary)
  • Audit token consistency and detect duplicates or unused variables
  • Generate theme variants (light/dark) and map semantic tokens to primitives
  • Create documentation and migration guides when refactoring tokens

Best practices

  • Prefer semantic names (e.g., color-primary) for application-facing tokens and descriptive names (e.g., blue-500) for primitives
  • Keep primitives and semantics separated and group tokens by category (colors, spacing, typography)
  • Preserve units from Figma and convert to target-platform formats when needed (px→rem for web)
  • Include Figma file references and generation timestamps for traceability
  • Validate token values and supply fallbacks for referenced tokens

Example use cases

  • Generate CSS custom properties for a web project and include Figma references in comments
  • Produce a TypeScript tokens module with export types for safe use in components
  • Create a Style Dictionary input JSON to build platform-specific artifacts (iOS/Android/web)
  • Produce a Tailwind config that maps design tokens to utility values
  • Audit a Figma file to list missing semantic tokens, unused variables, and naming inconsistencies

FAQ

What do I need open in Figma to run extraction?

You must have the target file open in Figma Desktop (MCP available) or provide a Figma file URL; the extractor reads variables via the desktop MCP API.

Which output formats are supported?

Supported formats include CSS custom properties, JavaScript/TypeScript objects, JSON, Sass/SCSS, Tailwind config, and Style Dictionary-compatible JSON.