MCP Chat Support System

# 🚀 Deploy to Hugging Face Spaces This guide will help you deploy the Gemini MCP Server to Hugging Face Spaces. ## Prerequisites 1. **Google Gemini API Key**: Get a free API key from [Google AI Studio](https://aistudio.google.com/app/apikey) 2. **Hugging Face Account**: Sign up at [huggingface.co](https://huggingface.co) 3. **GitHub Repository**: Your code should be in a GitHub repository ## Step-by-Step Deployment ### 1. Prepare Your Repository Make sure your repository contains: - `app.py` - Main FastAPI application - `requirements.txt` - Python dependencies - `app_config.py` - Configuration file - All other necessary files ### 2. Create a Hugging Face Space 1. Go to [Hugging Face Spaces](https://huggingface.co/spaces) 2. Click **"Create new Space"** 3. Choose settings: - **Owner**: Your username - **Space name**: `gemini-mcp-server` (or your preferred name) - **License**: Choose appropriate license - **SDK**: **Docker** - **Python version**: 3.9 - **Hardware**: CPU (free tier) ### 3. Connect Your Repository 1. In the Space creation form, select **"Repository"** as the source 2. Choose your GitHub repository 3. Set the **Repository path** to: `gemini-mcp-server/` (if your backend is in a subdirectory) 4. Click **"Create Space"** ### 4. Configure Environment Variables 1. Go to your Space's **Settings** tab 2. Scroll down to **"Repository secrets"** 3. Add the following secrets: ``` GEMINI_API_KEY=your_actual_gemini_api_key_here ``` **Optional environment variables:** ``` DATABASE_URL=sqlite:///./data/mcp_server.db ALLOWED_ORIGINS=* RATE_LIMIT_REQUESTS=100 RATE_LIMIT_PERIOD=3600 LOG_LEVEL=INFO ``` ### 5. Deploy 1. The Space will automatically start building 2. Monitor the build logs in the **"Logs"** tab 3. Once successful, your API will be available at: ``` https://your-username-gemini-mcp-server.hf.space ``` ## API Endpoints Once deployed, your API will have these endpoints: - `GET /` - Health check and server info - `GET /mcp/version` - MCP version information - `GET /mcp/capabilities` - Available capabilities - `POST /mcp/process` - Process single query - `POST /mcp/batch` - Process multiple queries - `GET /mcp/health` - Detailed health check ## Testing Your Deployment 1. **Health Check**: Visit your Space URL to see the welcome message 2. **API Test**: Use the `/mcp/health` endpoint to verify all services 3. **Integration**: Update your frontend to use the new backend URL ## Troubleshooting ### Common Issues: 1. **Build Fails**: Check the logs for missing dependencies 2. **API Key Error**: Ensure `GEMINI_API_KEY` is set correctly 3. **Port Issues**: The app automatically uses port 7860 for HF Spaces 4. **CORS Errors**: Check `ALLOWED_ORIGINS` configuration ### Getting Help: - Check the **Logs** tab in your Space for detailed error messages - Review the main `README.md` for more technical details - Open an issue in the repository if you encounter problems ## Cost - **Hugging Face Spaces**: Free tier available - **Google Gemini API**: Free tier with generous limits - **Total Cost**: $0 for basic usage ## Next Steps After successful deployment: 1. Update your frontend to point to the new backend URL 2. Test all functionality end-to-end 3. Monitor usage and performance 4. Consider upgrading to paid tiers if needed --- **🎉 Congratulations!** Your Gemini MCP Server is now deployed on Hugging Face Spaces!