Quick Start
As MCP Server (Recommended)
1. Add to your AI tool
Add imugi as an MCP server in your Claude Code or Cursor config:
{ "mcpServers": { "imugi": { "command": "npx", "args": ["-y", "imugi-ai", "mcp"] } }}2. Ask your AI to implement a design
Implement this login page design. Here's the reference: ./login-design.pngThe AI will use imugi’s tools to:
- Detect your project stack
- Start your dev server
- Capture screenshots
- Compare against the design
- Iterate until the code matches
As CLI
Generate code from a design
imugi generate ./design.png --output src/app/page.tsxCompare a design against a screenshot
imugi compare ./design.png --screenshot ./current.pngRun the full Boulder Loop
export ANTHROPIC_API_KEY=sk-ant-...imugiThen describe what you want:
> implement this design ./login-design.pngExport from Figma
# Export a Figma frameimugi figma "https://www.figma.com/design/FILE_KEY/name?node-id=42-1234"
# Export and compare against dev serverimugi figma "https://www.figma.com/design/FILE_KEY/name?node-id=42-1234" --compareRequires FIGMA_TOKEN environment variable.
Next Steps
- MCP Server Setup — Detailed MCP configuration
- Figma Integration — Connect Figma directly
- Configuration — Customize thresholds, viewports, and more