Figma Developer MCP (Model Context Protocol) is a Node.js server designed to bridge Figma design data with AI-powered coding agents, such as Cursor. Currently at version 0.11.0, the package sees a rapid release cadence, often with multiple updates within a month, indicating active development. Its core function is to fetch relevant metadata from the Figma API, then simplify and translate this data into a format optimized for AI models, focusing on layout and styling information. This targeted data reduction helps AI agents accurately "one-shot" designs into various frameworks, improving their precision compared to alternative methods like screenshot analysis. It differentiates itself by acting as a specialized intermediary, reducing context overload for AI models and facilitating direct integration of design specifications into code generation workflows via the Model Context Protocol.
npm install figma-developer-mcpVerified import paths — ran on the pinned version, not inferred.
This configuration JSON snippet demonstrates how to set up Framelink MCP for Figma within an AI client that supports Model Context Protocol servers, such as Cursor. It specifies the command to run the server and passes the Figma API key.
Review any custom server configurations or integrations for compatibility. The new stateless approach might require changes in how state is managed externally if needed.
If you are programmatically interacting with `getServerConfig()`, update your calls to remove the `isStdioMode` parameter and retrieve stdio mode from the returned `ServerConfig` object instead.
Review the project's documentation or GitHub repository for information on how to opt-out of telemetry if privacy is a concern.
If your environment relies on HTTP proxy auto-detection via environment variables (like `HTTP_PROXY`), verify that network requests from `figma-developer-mcp` are still routed correctly or reconfigure explicit proxy settings if necessary.
Ensure your Node.js environment is updated to version `20.20.0` or newer to run this package.
Ensure the `FIGMA_API_KEY` is correctly set in your AI client's configuration or as an environment variable, and that it's a valid personal access token with appropriate permissions. Check the Figma help documentation for creating and managing tokens.
Verify that the Figma link points to a valid and accessible node within the file. The node might have been deleted, moved, or the link itself is malformed.
Carefully review your AI client's MCP server configuration JSON against the provided quickstart example, ensuring all necessary fields are present and correctly formatted, especially `command` and `args`.