A powerful web-based AI coding assistant that allows users to edit GitHub repositories inside a browser. This application allows you to clone any GitHub repository, edit code in real-time, and get AI assistance with development tasks. It automatically deploys cloud containers on Railway with a universal Docker image that supports any Node.js, Python, or web project.

• 🤖 AI-Powered Code Assistant - Chat with Gemini AI to make code changes
• ☁️ Railway Cloud Containers - Each project runs in its own isolated cloud environment
• 🐳 Universal Docker Containers - Automatically detects and runs any Node.js, Python, or web project
• 📂 GitHub Repository Cloning - Clone and work with any public or private GitHub repo
• 🛠️ Live Development Tools - Edit files, run commands, restart servers in real-time
• 📡 Live Log Streaming - See your application logs in real-time as you develop
• 🎨 Modern UI - Built with Next.js, Mantine, and Tailwind CSS
• 🔄 Multi-step Agent - AI assistant can take multiple actions to complete complex tasks
• 👥 Multi-tenant Support - Multiple users can work on the same repository simultaneously

• Frontend: Next.js 15 with App Router, Mantine UI components
• AI Integration: Google Gemini via AI SDK
• Cloud Runtime: Railway containers with universal Docker images
• Container Registry: GitHub Container Registry (GHCR)
• Real-time Communication: Server-Sent Events for log streaming
• Styling: Tailwind CSS + Mantine component library

npm install

Create a .env.local file with the following variables:

# AI Configuration
GOOGLE_GENERATIVE_AI_API_KEY=your_google_ai_api_key_here

# Railway Configuration (server-side only)
RAILWAY_DEV_PROJECT_ID=your_railway_dev_project_id
RAILWAY_DEV_ENVIRONMENT_ID=your_railway_dev_environment_id
RAILWAY_TOKEN=your_railway_api_token

Follow the detailed setup guide in RAILWAY_SETUP.md to:

• Create a Railway project and get credentials
• Configure your Railway environment
• Set up the universal container deployment

npm run dev

Navigate to http://localhost:3000

To work with a GitHub repository:

1. Navigate to http://localhost:3000/your-project-name?repo=https://github.com/owner/repo
2. Add your GitHub token when requested (for private repos)
3. The system will automatically:
   • Deploy a new Railway container with universal Docker image
   • Generate a public Railway domain (e.g., project-user-production.up.railway.app)
   • Clone your repository inside the container
   • Auto-detect project type (Node.js, Python, etc.)
   • Install dependencies (npm install, pip install, etc.)
   • Start your development server with proper port configuration
   • Expose your app through path-based proxy routing
   • Stream real-time logs from both container and your app
   • Enable hot module replacement for instant changes

The AI assistant can help you with:

• File Operations:

  • read_file - View any file in your project
  • write_file - Create or modify files
  • list_files - Explore project structure

• Development Tools:

  • run_command - Execute terminal commands
  • restart_server - Restart your development server

• Real-time Features:

  • Live code changes with hot reloading
  • Real-time log streaming
  • Instant feedback on compilation errors

🌐 Live Demo: https://sr2e-starter-production.up.railway.app

This is a live example of a Next.js starter deployed through our system:

• User App: https://sr2e-starter-production.up.railway.app/ (Next.js with Mantine UI)
• Container API: https://sr2e-starter-production.up.railway.app/_container/health
• Live Logs: https://sr2e-starter-production.up.railway.app/_container/logs/stream

• "Show me the project structure and main components"
• "Add authentication to this Next.js app"
• "Fix the TypeScript errors in the components folder"
• "Add a new API endpoint for user management"
• "Optimize this React component for performance"
• "Set up a database connection and create user models"

Each deployed container uses a path-based proxy system:

• https://your-domain.up.railway.app/ - Your application (Next.js, React, etc.)
• https://your-domain.up.railway.app/about - All standard app routes
• https://your-domain.up.railway.app/_next/* - WebSocket HMR, static assets

• GET /_container/health - Container health check
• POST /_container/read_file - Read file contents
• POST /_container/write_file - Write/modify files
• POST /_container/list_files - List directory contents
• POST /_container/run_command - Execute commands
• POST /_container/restart_server - Restart the development server
• GET /_container/logs/stream - Real-time log streaming via Server-Sent Events

The proxy ensures zero conflicts between your app and AI agent tools while maintaining full HMR support.

# Development server with hot reloading
npm run dev

# Production build
npm run build

# Start production server
npm run start

# Lint code
npm run lint

# Testing scripts
npm run test:container      # Test container deployment
npm run test:files         # Test file operations
npm run test:logs          # Test log streaming
npm run test:all           # Run full test suite
npm run inspect:container  # Interactive container inspector

The system supports multiple users working on the same repository simultaneously:

• Each user gets a unique user ID stored in localStorage
• Railway services are named with format: {userId}-{repoName}
• Each user-repo combination gets its own isolated container
• Automatic Railway domain generation: {userId}-{repoName}.up.railway.app
• Real-time log streaming is per-user session
• No conflicts between users editing the same repository

The universal Docker container automatically detects and supports:

• Next.js Applications - npm run dev with HMR support
• React Applications - Create React App, Vite projects
• Express Servers - API servers and web applications
• Vue.js Applications - Nuxt.js and standard Vue projects
• General Node.js - Any project with package.json

• Flask Applications - Web applications and APIs
• Django Projects - Full-stack web applications
• FastAPI - Modern async API frameworks
• General Python - Any project with requirements.txt

• HTML/CSS/JS - Served via static file server
• Generated Sites - Gatsby, Hugo output directories

• Detects project type from package.json, requirements.txt
• Automatically installs dependencies
• Selects appropriate start command (npm run dev, python app.py, etc.)
• Configures ports for Railway deployment
• Handles both development and production modes

This application requires a modern browser with support for:

• ES2022 features
• Server-Sent Events (EventSource)
• Modern JavaScript APIs

1. Fork the repository
2. Create a feature branch
3. Test your changes with the provided testing scripts
4. Ensure all container functionality works
5. Submit a pull request

MIT License - see LICENSE file for details