f2c-mcp-server
# F2C MCP Server
[](https://smithery.ai/server/@f2c-ai/f2c-mcp)
[![npm version][npm-version-src]][npm-version-href]
[![npm downloads][npm-downloads-src]][npm-downloads-href]
[![github][github-src]][github-href]
[![node][node-src]][node-href]
[npm-version-src]: https://img.shields.io/npm/v/@f2c/mcp?style=flat&colorA=18181B&colorB=F0DB4F
[npm-version-href]: https://npmjs.com/package/@f2c/mcp
[npm-downloads-src]: https://img.shields.io/npm/dm/@f2c/mcp?style=flat&colorA=18181B&colorB=F0DB4F
[npm-downloads-href]: https://npmjs.com/package/@f2c/mcp
[github-src]: https://img.shields.io/badge/github-@f2c/mcp-blue?style=flat&colorA=18181B&colorB=F0DB4F
[github-href]: https://github.com/f2c-ai/f2c-mcp
[node-src]: https://img.shields.io/node/v/@f2c/mcp?style=flat&colorA=18181B&colorB=F0DB4F
[node-href]: https://nodejs.org/en/about/previous-releases
> Due to [Figma REST API rate limits](https://developers.figma.com/docs/rest-api/rate-limits/), if you are affected, please switch to [@f2c/mcp-plugin](https://www.npmjs.com/package/@f2c/mcp-plugin) for normal operation.
English | [简体中文](./README-zh-CN.md)
A Model Context Protocol server for Figma Design to Code using [F2C](https://f2c.yy.com/).
<a href="https://glama.ai/mcp/servers/@f2c-ai/f2c-mcp">
<img width="380" height="200" src="https://glama.ai/mcp/servers/@f2c-ai/f2c-mcp/badge" alt="f2c-mcp-server MCP server" />
</a>
## Features
<img alt="f2c" src="https://raw.githubusercontent.com/f2c-ai/f2c-mcp/main/docs/bannerv3.png" />
- 🎨 Pixel-Perfect HTML/CSS：F2C converts Figma designs to pixel-perfect HTML/CSS with precision.
- ⚛️ Multi-Framework Support：F2C generates React, CSS Modules, and Tailwind CSS code for fast development.
- 🧠 Figma Design Context：F2C integrates design context, ensuring compatibility with AI tools like Cursor.
- 🔗 Figma File URL Parsing：F2C converts design nodes via Figma URLs, streamlining workflows.
- 🖼️ Remote Image Localization：F2C automates downloading Figma images to local assets for efficiency.
## How it works
1. [Configure the Server](docs/en/GettingStarted.md) in an MCP-supported IDE (e.g., Cursor, Trae).
> recommended to use [Comate AI IDE](https://comate.baidu.com/zh/download/ai-ide)
2. Open your chat in IDE (e.g. agent mode in Cursor).
3. Paste a link to a Figma Node (Right-click any node in the Figma Layer panel to copy it).
4. Enter your requirements in the chat, such as fetching node data, downloading images, converting to code, etc.
## Configuration and Development
See [Configuration and Development](docs/en/GettingStarted.md)
## Data Privacy Notice
The logging tools integrated in this project are used solely for basic usage statistics and error log reporting. No sensitive information or user data is collected. All reported data is used exclusively to improve product quality and user experience.
## FAQ
See [FAQ](docs/en/FAQ.md)
## Credits
Thanks to:
+ [Framelink Figma MCP Server](https://github.com/GLips/Figma-Context-MCP) Give Cursor and other AI-powered coding tools access to your Figma files with this Model Context Protocol server.
+ [Cursor Talk to Figma MCP](https://github.com/sonnylazuardi/cursor-talk-to-figma-mcp) Allowing Cursor to communicate with Figma for reading designs and modifying them programmatically.
+ [Figma MCP Server](https://github.com/MatthewDailey/figma-mcp) This server provides tools for viewing, commenting, and analyzing Figma designs directly through the ModelContextProtocol.TDQS
Scored across 5 tools
The tools have some overlap in purpose, particularly between 'figma_get_file_data' and 'figma_get_file_meta', which could cause confusion about which to use for general file information. However, descriptions help clarify distinctions, such as 'figma_get_image_fills' for image resources versus 'figma_get_images' for node images, and 'figma_to_code' has a unique conversion function.
All tool names follow a consistent 'figma_' prefix with snake_case and descriptive verb_noun patterns, such as 'get_file_data' and 'to_code'. There are no deviations in naming conventions, making the set predictable and easy to understand.
With 5 tools, the count is reasonable for a Figma-to-code server, covering data retrieval, image handling, and conversion. It's slightly under the ideal range for comprehensive coverage but well-scoped for core functionalities without being overwhelming.
The tool set covers key operations like retrieving file data, metadata, images, and converting to code, but there are notable gaps, such as missing update or delete operations for Figma files, and no tools for managing design components or handling user interactions, which limits full lifecycle coverage in the design-to-code domain.