Sponsored by Deepsite.site

Tag

#IM

458 results found

Zapcode Figma Mcp

# Zapcode Figma MCP Server [![npm version](https://img.shields.io/npm/v/zapcode-figma-mcp.svg)](https://www.npmjs.com/package/zapcode-figma-mcp) Connect your Figma designs directly to AI tools like Claude Desktop, Claude Code, Cursor, and Cline. Get real-time HTML, CSS, images, and assets from selected Figma frames through the Model Context Protocol. ## Features - **Real-time Design Context** - Fetches live data from your currently selected Figma frame - **Complete Asset Export** - Supports SVG, PNG, JPG, PDF, and all Figma asset types - **Configurable Storage** - Save assets to custom paths or default workspace location - **Auto-organized Assets** - Assets grouped by type (svg/, png/, jpg/, etc.) - **Zero Configuration** - Works out of the box with npx - **Rich Context** - Provides HTML structure, CSS styles, images, and design specifications ## Prerequisites 1. **Node.js 16+** - [Download here](https://nodejs.org/) 2. Figma 3. **[Zapcode Figma Plugin](https://www.figma.com/community/plugin/1454956820198178710/zapcode)** - Install from Figma Community ## Installation ### Claude Desktop Add to your `claude_desktop_config.json`: **macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json` **Windows**: `%APPDATA%\Claude\claude_desktop_config.json` ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` Restart Claude Desktop after making changes. ### Claude Code Add to your `claude_code_config.json`: **macOS/Linux**: `~/.config/claude-code/claude_code_config.json` **Windows**: `%APPDATA%\claude-code\claude_code_config.json` ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### Cursor Add to your MCP settings configuration: ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### Cline (VS Code Extension) Add to Cline's MCP settings: ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### Windsurf Add to your Windsurf MCP configuration: ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### Zed Editor Add to your Zed settings (MCP support currently in preview): ```json { "context_servers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### Continue (VS Code/JetBrains) Add to Continue's configuration: ```json { "mcpServers": { "zapcode-figma": { "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ### GitHub Copilot Add to your GitHub Copilot MCP configuration: ```json { "servers": { "zapcode-figma": { "type": "stdio", "command": "npx", "args": ["-y", "zapcode-figma-mcp"] } } } ``` ## How It Works 1. **Install Plugin** - Add Zapcode plugin to Figma Desktop 2. **Run Plugin** - Launch Zapcode plugin from Figma 3. **Select Frame** - Choose any frame in your Figma design 4. **Ask AI** - Ask AI to bring you the figma context 5. **Get Context** - Receive complete design data and assets ## Troubleshooting ### "Figma plugin is not connected" **Cause**: MCP server cannot reach the Zapcode Figma plugin. **Solution**: 1. Open Figma Desktop App 2. Open your design file 3. Run Zapcode plugin (Plugins → Zapcode) 4. Ensure plugin window stays open 5. Retry connecting using toggle 5. Try the MCP tool again ### Assets Not Saving **Cause**: Permission issues or invalid path. **Solution**: 1. Ensure the assets path exists or parent directory is writable 2. Use absolute paths for clarity 3. Check file permissions on the target directory ## Asset Organization Assets are automatically organized by type: ``` assets/ ├── svg/ │ ├── icon_component.svg │ └── logo.svg ├── png/ │ ├── screenshot_1.png │ └── image_export.png └── jpg/ └── photo.jpg ``` ## Supported Clients This server works with any MCP-compatible client, including: - [Claude Desktop](https://modelcontextprotocol.io/clients#claude-desktop) - Anthropic's desktop application - [Claude Code](https://claude.com/claude-code) - Anthropic's official CLI for Claude - [Cursor](https://www.cursor.com/) - AI-first code editor - [Cline](https://github.com/cline/cline) - Autonomous coding agent for VS Code - [Windsurf](https://codeium.com/windsurf) - AI-powered IDE with MCP Plugin Store - [Zed](https://zed.dev/) - High-performance code editor (MCP in preview) - [Continue](https://continue.dev/) - Open-source IDE extension - [GitHub Copilot](https://github.com/features/copilot) - With MCP support in VS Code ## Links - [NPM Package](https://www.npmjs.com/package/zapcode-figma-mcp) - [Zapcode Figma Plugin](https://www.figma.com/community/plugin/1454956820198178710/zapcode) - [Model Context Protocol](https://modelcontextprotocol.io/) - [MCP Servers Collection](https://github.com/modelcontextprotocol/servers) --- **Sources:** - [Model Context Protocol Official Servers](https://github.com/modelcontextprotocol/servers) - [GitHub MCP Server](https://github.com/github/github-mcp-server) - [MCP Clients Directory](https://github.com/punkpeye/awesome-mcp-clients) - [Official MCP Client Examples](https://modelcontextprotocol.io/clients)

Vision Mcp Server | 图片分析 Mcp

This MCP addresses the visual recognition limitations of text-based models by enabling accurate image description and identification, making it excellent for AI-assisted reference design interface analysis. It currently supports dropping links into the dialog box or placing images in the project folder for recognition. The tool can be integrated with MCP platforms like Claude Code, Cline, and Trae. Beyond programming applications, it also provides visual recognition capabilities for models that lack native image processing functionality. For visual models, users can select their preferred model from ModelScope community and replace it during MCP configuration setup. 📱 Daily Use Cases: Send screenshots to directly identify errors or issues Share image links or place screenshots in the project folder for AI-assisted layout optimization Submit product image links to generate promotional copy 该mcp可以解决文字模型图片识别的视觉的问题,可以准确识别描述图片,用来给AI看参考设计界面很nice~ 目前支持丢链接到对话框,以及把图片放到项目文件夹进行识别。 支持加入到Claude Code,Cline和Trae等mcp工具中。 除了编程外,如果你使用的模型本身不支持视觉图片识别,也可以使用~ 视觉模型可以自己去魔搭社区选一个自己喜欢的,在填写mcp配置的时候替换即可 📱 日常使用场景 - 截图发过去,直接告诉哪里出错了 - 丢过去一个图片链接或者截图放到项目文件夹内,让AI帮忙优化布局 - 发个产品图链接,让AI写推广文案