@ramme-io/create-app

The official CLI tool to bootstrap a Ramme application.

One command. Zero configuration. A production-grade project in seconds.

Quick Start

No global installation required. Run directly with npx:

npx @ramme-io/create-app@latest my-project

Or using the npm create shorthand:

npm create @ramme-io/app@latest my-project

Then start developing:

cd my-project
pnpm install
pnpm run dev

Your app is running at http://localhost:5173.

What's Included

The CLI generates a fully configured project with:

• Vite — Instant dev server start and sub-second HMR.
• React 19 — The standard library for composable user interfaces.
• TypeScript — Static typing for scalable, robust application logic.
• Tailwind CSS — Utility-first styling pre-configured with Ramme design tokens.
• @ramme-io/ui — The full Ramme component library, pre-installed and ready to import.
• @ramme-io/kernel — The core routing and rendering engine, pre-wired to the manifest system.

Generated Project Structure

my-project/
├── src/
│   ├── config/
│   │   ├── app.manifest.ts    # App title, description, theme settings
│   │   └── sitemap.ts         # Page definitions and navigation structure
│   ├── core/
│   │   └── component-registry.tsx  # Maps component names to React components
│   ├── pages/                 # Your page components
│   └── data/                  # Mock data and seed files
├── tailwind.config.js         # Pre-loaded with @ramme-io/ui preset
├── vite.config.ts
└── package.json

🤖 AI Integration

The starter template includes the Universal AI Gateway with support for 6 providers and token-by-token streaming:

• ✅ Providers: Gemini, OpenAI, Anthropic, Grok, DeepSeek, Ollama
• ✅ Streaming: Real-time text generation with SSE
• ✅ Secure: API keys stay server-side
• ✅ Pre-configured: Proxy endpoint and client utilities included

Quick Setup

1. Create a .env file:

   VITE_DEFAULT_GEMINI_KEY=your-api-key-here

2. Use in your components:

   import { streamAiText } from './services/aiProxyClient';
      
   await streamAiText(
     { prompt: 'Hello, AI!' },
     (chunk) => console.log(chunk)  // Text streams token-by-token!
   );

📚 Full AI Gateway Setup Guide →

Related Packages

• @ramme-io/ui — The component library.
• @ramme-io/kernel — The core routing and rendering engine.

Documentation

For full documentation, visit https://ramme.io/docs.

License

MIT License — Copyright © 2026 Ramme Framework.

See LICENSE for details.