Figsor
Bridges Cursor AI to Figma for chat-driven design creation and editing directly on the Figma canvas. It provides over 45 tools to manage layers, layouts, components, and variables through natural language commands.
README
Figsor
Chat in Cursor. Design in Figma.
Figsor is an MCP server that bridges Cursor AI to Figma, enabling chat-driven design creation and editing — directly on your Figma canvas.
Cursor → MCP (stdio) → Figsor Server → WebSocket → Figma Plugin → Design on Canvas
Setup
1. Install the Figma Plugin (Sideload)
Clone this repo and import the plugin into Figma:
https://github.com/AsifKabirAntu/figsor.git
In Figma: Plugins → Development → Import plugin from manifest → select figsor/figma-plugin/manifest.json
2. Add to Cursor
Open your Cursor MCP settings and add:
{
"mcpServers": {
"figsor": {
"command": "npx",
"args": ["-y", "figsor"]
}
}
}
3. Start Designing
- Open a Figma file
- Run the Figsor plugin (Plugins → Development → Figsor)
- Chat in Cursor:
"Create a mobile login screen with email and password fields"
"Design a dashboard with a sidebar, KPI cards, and charts"
"Edit the selected frame — make the button rounded and change the color to blue"
Available Tools (45+)
Create & Layout
| Tool | Description |
|---|---|
create_frame |
Create frames (screens, sections, cards) |
create_text |
Add text with font, size, weight, color |
create_rectangle |
Create rectangles and shapes |
create_ellipse |
Create circles and ovals |
create_line |
Create lines and dividers |
create_svg_node |
Create icons and vector graphics from SVG |
set_auto_layout |
Configure flexbox-style auto-layout |
modify_node |
Edit any existing element |
set_stroke |
Add borders and strokes |
set_effects |
Add shadows and blur effects |
delete_node |
Remove elements |
move_to_parent |
Restructure the layer hierarchy |
Read & Inspect
| Tool | Description |
|---|---|
get_selection |
Read the current selection |
get_page_structure |
Get the full page tree |
read_node_properties |
Inspect any node's properties |
find_nodes |
Search for elements by name or type |
set_selection |
Select and zoom to elements |
get_local_styles |
Read the file's design tokens |
list_components |
Browse available components |
create_component_instance |
Use existing components |
detach_instance |
Convert instances to frames |
Vector Drawing & Advanced Fills
| Tool | Description |
|---|---|
create_vector |
Draw custom shapes with the pen tool |
boolean_operation |
Union, subtract, intersect, or exclude shapes |
flatten_nodes |
Flatten nodes into a single editable vector |
set_fill |
Apply solid colors, linear/radial/angular/diamond gradients, multiple fills |
Image, Typography & Constraints
| Tool | Description |
|---|---|
set_image_fill |
Place image fills on nodes |
style_text_range |
Apply mixed styling within text |
set_constraints |
Set responsive constraints |
list_available_fonts |
Discover available fonts |
Component & Variable Tools
| Tool | Description |
|---|---|
create_component |
Create a new main component |
create_component_set |
Combine components into a variant set |
create_variable_collection |
Create a design token collection with modes |
create_variable |
Create a COLOR, FLOAT, STRING, or BOOLEAN token |
bind_variable |
Bind a token to a node property |
get_variables |
List all variable collections and tokens |
SVG Export & Animation
| Tool | Description |
|---|---|
export_as_svg |
Export any node as SVG markup |
show_animation_preview |
Live animated SVG previews + ZIP download |
AI-Powered SVG (Quiver)
| Tool | Description |
|---|---|
quiver_generate_svg |
Generate SVG graphics from text prompts |
quiver_vectorize_svg |
Convert raster images to clean SVG |
Peer Design (Multi-Agent)
| Tool | Description |
|---|---|
spawn_design_agent |
Spawn AI designer agents with visible cursors |
dismiss_design_agent |
Remove an agent cursor |
dismiss_all_agents |
Remove all agent cursors |
Design Craft Guide
| Tool | Description |
|---|---|
get_design_craft_guide |
Professional design rules — typography, color, spacing, anti-AI-slop |
Pro: Design System Integration
Connect your Figma design system libraries so the AI uses YOUR components, not generic ones.
- Scan & import library components
- Search across your design system
- Save & switch between libraries
- Generate designs with your DS
Get Figsor Pro → — $9 one-time purchase
Requirements
- Node.js 18 or later
- Figma desktop or web app
- Cursor IDE with MCP support (or any MCP-compatible client — see below)
Using with Claude Code
Figsor works with any MCP client, not just Cursor. To use it with Claude Code:
claude mcp add --transport stdio --scope project figsor -- npx -y figsor
Or add to .mcp.json in your project root:
{
"mcpServers": {
"figsor": {
"type": "stdio",
"command": "npx",
"args": ["-y", "figsor"]
}
}
}
Then follow the same steps — open Figma, run the Figsor plugin, and chat.
Configuration
| Environment Variable | Default | Description |
|---|---|---|
FIGSOR_PORT |
3055 |
WebSocket server port |
License
MIT © Asif Kabir
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。