Figma MCP Server
Brings Figma designs directly into MCP-compatible environments to view specifications and export frames as high-resolution images. It features smart section splitting for large frames and integrates with Jira to automatically fetch designs from ticket links.
README
Figma MCP Server
A Model Context Protocol (MCP) server that brings Figma designs directly into Claude Code. View design specifications, export frames as images, and reference visual context without leaving your development environment.
Overview
Implementing designs accurately requires constant reference to Figma. This MCP server eliminates context switching by:
- Fetching design information - Get file names, frame details, and dimensions
- Exporting images - Download frames as high-resolution PNGs
- Smart section splitting - Large frames are automatically split into readable sections
- Seamless integration - Works standalone or integrated with jira-mcp
Features
| Feature | Description |
|---|---|
| Design Info | File name, last modified, frame dimensions |
| Image Export | Export frames as PNG, SVG, JPG, or PDF |
| Auto-Scaling | 2x scale by default for retina clarity |
| Section Detection | Large frames split into individual sections |
| Batch Export | Export multiple frames efficiently |
| URL Parsing | Supports file, design, and prototype URLs |
Installation
Prerequisites
- Node.js 18+
- Claude Code CLI
- Figma account with API access
Step 1: Add to Claude Code
claude mcp add --transport stdio figma -- npx -y @rui.branco/figma-mcp
Step 2: Get Your Figma Token
- Go to Figma Account Settings
- Scroll down to "Personal access tokens"
- Click "Generate new token"
- Enter a description (e.g., "Claude Code MCP")
- Click "Generate token"
- Copy the token (you won't be able to see it again)
Step 3: Configure Credentials
Run the setup with your token:
npx @rui.branco/figma-mcp setup "YOUR_FIGMA_TOKEN"
Or run interactively (will prompt for the token):
npx @rui.branco/figma-mcp setup
Step 4: Verify
Restart Claude Code and run /mcp to verify the server is connected.
Alternative: Manual Installation
If you prefer to install manually:
git clone https://github.com/rui-branco/figma-mcp.git ~/.config/figma-mcp
cd ~/.config/figma-mcp && npm install
node setup.js
Then add to Claude Code:
claude mcp add --transport stdio figma -- node $HOME/.config/figma-mcp/index.js
Usage
Fetch a Design
> Get this Figma design: https://www.figma.com/design/ABC123/MyProject?node-id=1-234
Example Output
# Figma File: MyProject
**Last Modified:** 2025-01-15T10:30:00Z
## Selected Frame: Login Screen
**Type:** FRAME
**Size:** 1440 x 900
### Sections (4):
- **Header** (FRAME, 1440x80)
- **Login Form** (FRAME, 400x350)
- **Footer** (FRAME, 1440x60)
- **Background** (FRAME, 1440x900)
### Exported Sections:
- Header: ~/.config/figma-mcp/exports/ABC123_1_234.png
- Login Form: ~/.config/figma-mcp/exports/ABC123_1_235.png
- Footer: ~/.config/figma-mcp/exports/ABC123_1_236.png
[Images displayed inline]
Smart Section Export
For large frames (>1500px wide or >2000px tall), the server automatically:
- Detects child frames, components, and groups
- Exports each section separately at 2x scale
- Provides better detail than a single compressed image
- Limits to 10 sections by default (configurable)
This is especially useful for:
- Full-page designs with multiple sections
- Component libraries
- Design system documentation
API Reference
Tools
| Tool | Description | Parameters |
|---|---|---|
figma_get_design |
Fetch design from URL | url (required), exportImage, exportChildren, maxChildren, scale |
figma_export_frame |
Export specific frame | fileKey, nodeId (required), format, scale |
Parameters
| Parameter | Default | Description |
|---|---|---|
exportImage |
true |
Export images |
exportChildren |
true |
Split large frames into sections |
maxChildren |
10 |
Maximum sections to export |
scale |
2 |
Export scale (1-4) |
format |
png |
Export format (png, svg, jpg, pdf) |
Configuration
Config stored at ~/.config/figma-mcp/config.json:
{
"token": "YOUR_FIGMA_TOKEN"
}
Error Handling
The server provides clear error messages:
| Error | Meaning | Solution |
|---|---|---|
Rate limit exceeded |
Too many API requests | Wait a few minutes |
Access denied |
Invalid token or no file access | Check token permissions |
File not found |
Invalid URL or deleted file | Verify the Figma URL |
Integration with Jira MCP
When used alongside jira-mcp:
- Figma links in Jira tickets are automatically detected
- Designs are fetched without manual intervention
- Images appear inline with ticket context
This creates a seamless workflow:
> Get ticket PROJ-123
# Returns ticket details + auto-fetched Figma designs
Security
- Tokens stored locally in
~/.config/figma-mcp/config.json - Config excluded from git via
.gitignore - Tokens only transmitted to Figma API
- Exports saved to
~/.config/figma-mcp/exports/
License
MIT
Related
- jira-mcp - Jira MCP server with Figma integration
- Model Context Protocol - MCP specification
- Figma API - Figma REST API documentation
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。