Fusion 360 MCP

# 🚀 START HERE - Quick Setup Guide ## ✅ Setup Complete! Everything is configured and ready to use with Ollama (your local models). --- ## 📋 How to Run FusionMCP ### **Step 1: Start Ollama Server** Open **Terminal 1**: ```bash ollama serve ``` Leave this running. You should see: ``` Ollama is running ``` --- ### **Step 2: Start MCP Server** Open **Terminal 2**: ```bash cd ~/Desktop/fusion360-mcp source venv/bin/activate python -m mcp_server.server ``` You should see: ``` ✅ Logger initialized with level INFO ✅ System prompt loaded ✅ Initialized MCP Router with providers: ['ollama'] ✅ MCP Server started on 127.0.0.1:9000 ``` Leave this running! --- ### **Step 3: Open Fusion 360** 1. **Launch Fusion 360** 2. **Load the Add-in**: - Press **Shift + S** (or Tools → Add-Ins → Scripts and Add-Ins) - Click **Add-Ins** tab - Find **FusionMCP** in the list - Select it and click **Run** 3. **You should see**: "FusionMCP Add-in Started!" 4. **Optional**: Check "Run on Startup" to auto-load next time --- ### **Step 4: Use MCP Assistant** 1. Look for **MCP Assistant** button in Fusion 360 toolbar 2. Click it to open the assistant 3. **Try these commands**: ``` Create a 20mm cube ``` ``` Make a cylindrical shaft 10mm diameter and 50mm long ``` ``` Design a mounting bracket 100x50mm with 4 M5 holes at corners ``` 4. Watch the magic happen! 🎉 --- ## 🎯 Your Configuration **Model**: llama3.1:latest (your best local model) **Fallback**: phi4-mini (if llama3.1 fails) **Server**: http://localhost:9000 **Mode**: 100% Local & Offline --- ## 🐛 Troubleshooting ### Issue: "Connection refused" ```bash # Make sure MCP server is running curl http://127.0.0.1:9000/health # Should return: {"status":"healthy","providers":["ollama"]...} ``` ### Issue: "Ollama not found" ```bash # Check Ollama is running curl http://localhost:11434/api/tags # Should return list of models ``` ### Issue: Add-in not visible in Fusion 1. Restart Fusion 360 2. Check folder name is exactly `FusionMCP` 3. Go to Tools → Add-Ins → Scripts and Add-Ins → Add-Ins tab ### Issue: Slow responses ```bash # Use faster model - edit config.json: "default_model": "ollama:phi4-mini" ``` --- ## 📊 Check Server Status ```bash # Health check curl http://127.0.0.1:9000/health # List available models curl http://127.0.0.1:9000/models # View conversation history curl http://127.0.0.1:9000/history?limit=5 ``` --- ## 📚 Documentation - **Full Setup Guide**: `SETUP_OLLAMA.md` - **Quick Start**: `QUICKSTART.md` - **Complete Docs**: `README.md` - **Architecture**: `ARCHITECTURE.md` --- ## 🎨 Your Available Models Based on your Ollama installation: - **llama3.1** - Best quality (default) - **phi4-mini** - Fast & small - **gemma3** - Good balance - **qwen3-coder** - Code-focused To switch models, edit `config.json`: ```json "default_model": "ollama:phi4-mini" ``` --- ## ✅ Quick Checklist Before testing: - [ ] Ollama running (`ollama serve`) - [ ] MCP server running (`python -m mcp_server.server`) - [ ] Fusion 360 open - [ ] FusionMCP add-in loaded - [ ] MCP Assistant button visible --- ## 🎉 You're Ready! Everything is configured to use **your new FusionMCP system** with: - ✅ Multi-model architecture - ✅ Intelligent fallback chain - ✅ Complete logging & caching - ✅ Production-ready code - ✅ Full test suite **Next**: Open Fusion 360 and start designing with AI! 🚀 --- **Questions?** Check the docs or run tests: ```bash ./run_tests.sh ```