GoHighLevel MCP Server

# 🚀 Cloud Deployment Guide - ChatGPT Integration ## 🎯 Overview To connect your GoHighLevel MCP Server to ChatGPT, you need to deploy it to a **publicly accessible URL**. Here are the best options: --- ## 🌟 **Option 1: Railway (Recommended - Free Tier)** ### **Why Railway?** - ✅ Free tier available - ✅ Automatic HTTPS - ✅ Easy GitHub integration - ✅ Fast deployment ### **Deployment Steps:** 1. **Sign up at [Railway.app](https://railway.app)** 2. **Create New Project from GitHub:** - Connect your GitHub account - Import this repository - Railway will auto-detect the Node.js app 3. **Set Environment Variables:** ``` GHL_API_KEY=your_api_key_here GHL_BASE_URL=https://services.leadconnectorhq.com GHL_LOCATION_ID=your_location_id_here NODE_ENV=production PORT=8000 ``` 4. **Deploy:** - Railway will automatically build and deploy - You'll get a URL like: `https://your-app-name.railway.app` 5. **For ChatGPT Integration:** ``` MCP Server URL: https://your-app-name.railway.app/sse ``` --- ## 🌟 **Option 2: Render (Free Tier)** ### **Deployment Steps:** 1. **Sign up at [Render.com](https://render.com)** 2. **Create Web Service:** - Connect GitHub repository - Select "Web Service" - Runtime: Node 3. **Configuration:** ``` Build Command: npm run build Start Command: npm start ``` 4. **Environment Variables:** (Same as above) 5. **For ChatGPT:** ``` MCP Server URL: https://your-app-name.onrender.com/sse ``` --- ## 🌟 **Option 3: Vercel (Free Tier)** ### **Deploy with One Click:** 1. **Click Deploy Button:** [](https://vercel.com/new/clone?repository-url=https://github.com/your-username/ghl-mcp-server) 2. **Add Environment Variables** during setup 3. **For ChatGPT:** ``` MCP Server URL: https://your-app-name.vercel.app/sse ``` --- ## 🌟 **Option 4: Heroku (Paid)** ### **Deployment Steps:** 1. **Install Heroku CLI** 2. **Deploy Commands:** ```bash heroku create your-app-name heroku config:set GHL_API_KEY=your_key_here heroku config:set GHL_BASE_URL=https://services.leadconnectorhq.com heroku config:set GHL_LOCATION_ID=your_location_id_here heroku config:set NODE_ENV=production git push heroku main ``` 3. **For ChatGPT:** ``` MCP Server URL: https://your-app-name.herokuapp.com/sse ``` --- ## 🎯 **Quick Test Your Deployment** Once deployed, test these endpoints: ### **Health Check:** ``` GET https://your-domain.com/health ``` Should return: ```json { "status": "healthy", "server": "ghl-mcp-server", "tools": { "total": 21 } } ``` ### **Tools List:** ``` GET https://your-domain.com/tools ``` Should return all 21 MCP tools. ### **SSE Endpoint (for ChatGPT):** ``` GET https://your-domain.com/sse ``` Should establish Server-Sent Events connection. --- ## 🔗 **Connect to ChatGPT** ### **Once your server is deployed:** 1. **Open ChatGPT Desktop App** 2. **Go to:** Settings → Beta Features → Model Context Protocol 3. **Add New Connector:** - **Name:** `GoHighLevel MCP` - **Description:** `Connect to GoHighLevel CRM` - **MCP Server URL:** `https://your-domain.com/sse` - **Authentication:** `OAuth` (or None if no auth needed) 4. **Save and Connect** ### **Test the Connection:** Try asking ChatGPT: ``` "List all available GoHighLevel tools" "Create a contact named Test User with email test@example.com" "Show me recent conversations in GoHighLevel" ``` --- ## 🚨 **Troubleshooting** ### **Common Issues:** 1. **502 Bad Gateway:** - Check environment variables are set - Verify GHL API key is valid - Check server logs for errors 2. **CORS Errors:** - Server includes CORS headers for ChatGPT - Ensure your domain is accessible 3. **Connection Timeout:** - Free tier platforms may have cold starts - First request might be slow 4. **SSE Connection Issues:** - Verify `/sse` endpoint is accessible - Check browser network tab for errors ### **Debug Commands:** ```bash # Check server status curl https://your-domain.com/health # Test tools endpoint curl https://your-domain.com/tools # Check SSE connection curl -H "Accept: text/event-stream" https://your-domain.com/sse ``` --- ## 🎉 **Success Indicators** ### **✅ Deployment Successful When:** - Health check returns `status: "healthy"` - Tools endpoint shows 21 tools - SSE endpoint establishes connection - ChatGPT can discover and use tools ### **🎯 Ready for Production:** - All environment variables configured - HTTPS enabled (automatic on most platforms) - Server responding to all endpoints - ChatGPT integration working --- ## 🔐 **Security Notes** - ✅ All platforms provide HTTPS automatically - ✅ Environment variables are encrypted - ✅ No sensitive data in code repository - ✅ CORS configured for ChatGPT domains only --- ## 💰 **Cost Comparison** | Platform | Free Tier | Paid Plans | HTTPS | Custom Domain | |----------|-----------|------------|-------|---------------| | **Railway** | 512MB RAM, $5 credit | $5/month | ✅ | ✅ | | **Render** | 512MB RAM | $7/month | ✅ | ✅ | | **Vercel** | Unlimited | $20/month | ✅ | ✅ | | **Heroku** | None | $7/month | ✅ | ✅ | **Recommendation:** Start with Railway's free tier! --- ## 🚀 **Next Steps** 1. **Choose a platform** (Railway recommended) 2. **Deploy your server** following the guide above 3. **Test the endpoints** to verify everything works 4. **Connect to ChatGPT** using your new server URL 5. **Start managing GoHighLevel through ChatGPT!** Your GoHighLevel MCP Server will be accessible at: ``` https://your-domain.com/sse ``` **Ready to transform ChatGPT into your GoHighLevel control center!** 🎯