UICanvas
🎨 UI preview canvas for developers — see AI-generated designs instantly, no Figma needed.
开发者的 UI 预览画布 — 即时查看 AI 生成的设计效果,内置组件素材库,无需专业设计工具。
Power your agent
การติดตั้งจากผู้เผยแพร่ ไม่ใช่เอนด์พอยต์ของ PlugMCP
URL แบบรีโมตหรือคำสั่งบนเครื่องของคุณ? การติดตั้งสองแบบต่างกันอย่างไร
รันบน PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
เปิดYou're a developer, not a designer. But you still need to see what the AI is generating.
UICanvas gives you an infinite canvas right in your browser. Connect it to any AI agent via MCP, and watch your UI designs appear in real-time — with a built-in component library to get started fast.
git clone https://github.com/markvely/uicanvas.git
cd uicanvas
npm install
node server.js
Open http://localhost:3200 — you'll see the canvas with demo artboards and the component library.
node server.js --stdio
Add to your MCP client config (Cursor, Claude Desktop, etc.):
{
"mcpServers": {
"uicanvas": {
"command": "node",
"args": ["/path/to/uicanvas/server.js", "--stdio"]
}
}
}
Now your AI agent can design UI directly on the canvas! 🎉
If you are an AI Coding Agent reading this documentation to understand how to use UICanvas, you MUST explicitly obey the following boundaries:
localhost:3200.init_project, create_artboard, write_html, etc.).mcp_uicanvas_open_canvas). DO NOT look for an exact name match like open_canvas. Instead, perform a substring match (open_canvas) on your available tools. If they are truly missing despite substrings, STOP IMMEDIATELY. Tell the user: "I cannot find the UICanvas MCP tools. Please Reload your VSCode Window or start a new Agent chat session so the tools can mount."create_artboard and write_html. You are strictly forbidden from writing .html files to the user's local disk or just spitting out code blocks in the chat. Render it in the canvas!Pre-flight Context Check & Professional Designer Mindset: Before you output any design code, you MUST:
package.json, README.md, or the main app files to determine the project type, human language (e.g., English, Chinese), tech stack, and expected visual style.AI Agent ──MCP stdio──▶ Node.js Server ──WebSocket──▶ Browser Canvas
▲ │
└────── DOM query ◀──────────┘
The AI agent sends design commands via MCP → the server forwards them via WebSocket → the browser renders HTML on the infinite canvas → you see the result instantly.
No need to design from scratch. UICanvas ships with professional components:
uicanvas/
├── server.js # Express + WS + MCP entry
├── lib/
│ ├── mcp-tools.js # MCP tool definitions
│ ├── ws-bridge.js # WebSocket bridge
│ └── artboard-manager.js # Artboard state management
├── public/
│ ├── index.html # Canvas UI + demo
│ ├── css/ # Dark theme + design tokens
│ └── js/ # Canvas engine + artboard renderer
└── components/
├── index.js # Gallery renderer
├── ios/ # iOS components (5 modules)
└── web/ # Web components (3 modules)
Contributions welcome! Fork → Branch → PR.
UICanvas 是一个面向开发者的 UI 预览画布。在 Web 开发过程中,让 AI 代理(如 Cursor、Copilot、Claude)直接在画布上生成和预览 UI 设计。
npm install && node server.js,浏览器打开即用UI preview canvas for developers. See AI-generated designs instantly with built-in component library. No Figma needed.
เครื่องมือ
ยังไม่มีรายการเครื่องมือ เซิร์ฟเวอร์แบบรีโมตจะได้รายการหลังการตรวจสอบสถานะครั้งแรก
FAQ
คัดลอกคอนฟิก stdio สำหรับ Cursor, Claude Desktop, Claude Code หรือ VS Code ได้ในหน้านี้ คอนฟิกนี้ใช้การติดตั้งของผู้เผยแพร่ ไม่ใช่พร็อกซีของ PlugMCP
ระเบียนนี้ไม่ได้ระบุตัวแปรสภาพแวดล้อมที่จำเป็น
Uicanvas ทำงานแบบโลคัลผ่าน stdio
ตรวจสอบล่าสุดเมื่อ 2026-10-26 UI preview canvas for developers. See AI-generated designs instantly with built-in component library. No Figma needed.