Figma MCP Server

A Model Context Protocol (MCP) server that connects to Figma's API, allowing AI tools and LLMs to access and work with your Figma designs.

Features

• Design Data Extraction: Extract components, styles, and text from your Figma designs
• Design System Analysis: Analyze design system consistency and patterns
• UI Content Management: Extract and organize all UI copy from designs
• Development Handoff: Generate comprehensive documentation for developers
• Seamless AI Integration: Connect your designs to AI tools like Claude, Cursor, and other MCP-compatible clients

Getting Started

Prerequisites

• Node.js 16 or higher
• Figma Personal Access Token (Get it from your Figma account settings)

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/figma-mcp-server.git
   cd figma-mcp-server
   

2. Install dependencies:

   npm install
   

3. Create a .env file in the project root:

   FIGMA_API_TOKEN=your_figma_personal_access_token
   API_KEY=your_secure_api_key
   TRANSPORT_TYPE=stdio
   

4. Build the server:

   npm run build
   

5. Start the server:

   npm start
   

Connecting to Clients

Claude for Desktop

1. Open or create the Claude for Desktop configuration file:

   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json

2. Add the following configuration:

   {
     "mcpServers": {
       "figma": {
         "command": "node",
         "args": ["/absolute/path/to/figma-mcp-server/build/index.js"],
         "env": {
           "FIGMA_API_TOKEN": "your_figma_personal_access_token",
           "TRANSPORT_TYPE": "stdio"
         }
       }
     }
   }
   

3. Restart Claude for Desktop

Cursor

Global Configuration

Create or edit Cursor's MCP configuration file:

• macOS: ~/Library/Application Support/Cursor/mcp.json
• Windows: %APPDATA%\Cursor\mcp.json

{
  "mcpServers": {
    "figma-mcp": {
      "url": "http://localhost:3000/sse",
      "env": {
        "API_KEY": "your_secure_api_key"
      }
    }
  }
}

Project-Specific Configuration

1. Create a .cursor directory in your project root:

   mkdir -p .cursor
   

2. Create an mcp.json file inside that directory:

   {
     "mcpServers": {
       "figma-mcp": {
         "url": "http://localhost:3000/sse",
         "env": {
           "API_KEY": "your_secure_api_key"
         }
       }
     }
   }
   

Available Tools

Tool	Description
get-file-info	Get basic information about a Figma file
get-nodes	Get specific nodes from a Figma file
get-components	Get component information from a Figma file
get-styles	Get style information from a Figma file
get-comments	Get comments from a Figma file
search-file	Search for elements in a Figma file by type, name, etc.
extract-text	Extract all text elements from a Figma file

Available Prompts

• analyze-design-system - Analyze design system components and styles for consistency
• extract-ui-copy - Extract and organize all UI copy from designs
• generate-dev-handoff - Generate development handoff documentation based on designs

Usage Examples

Using with Claude:

Can you analyze the design system in my Figma file with key abc123? Look for consistency in color usage and typography.

Using with Cursor:

Generate React components for the buttons from my Figma file with key abc123, using tailwind CSS.

Environment Variables

Variable	Description	Default
FIGMA_API_TOKEN	Your Figma Personal Access Token	(Required)
API_KEY	Security key for API authentication	(Required)
TRANSPORT_TYPE	Transport method (stdio or sse)	stdio
PORT	Port for SSE transport	3000

Architecture

This MCP server:

1. Connects to the Figma API using your personal access token
2. Exposes a standardized interface following the Model Context Protocol
3. Provides tools, resources, and prompts that LLMs can use to interact with your Figma designs
4. Supports both stdio transport (local connections) and SSE transport (remote connections)

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.