Skip to main content
Glama
MCP_FIXES.md1.69 kB
# MCP Server Fixes Applied ## Issues Fixed 1. **Python Path**: Changed from `python` to full path `C:\AIStack-MCP\.venv\Scripts\python.exe` - Ensures the venv with all dependencies is used - Avoids PATH issues 2. **Environment Variables**: Added `PYTHONPATH=C:\AIStack-MCP` - Ensures Python can find the server module 3. **Configuration**: Updated `.cursor/mcp_settings.json` and copied to Cursor config directory - Location: `%APPDATA%\Cursor\User\mcp_settings.json` ## Verification All components verified: - ✅ Server file exists - ✅ Venv Python exists - ✅ Config file in Cursor directory - ✅ Server starts successfully - ✅ Tools registered (5 tools: semantic_search, analyze_patterns, get_context, generate_code, index_workspace) ## Next Steps **IMPORTANT: Restart Cursor completely** for the MCP server to load. After restart: 1. Open Cursor 2. Open any project (or C:\AIStack) 3. The `code-intelligence` MCP server should be available 4. Try: "Use code-intelligence to index_workspace with file_extensions .py,.js,.ts and max_files 100" ## Troubleshooting If tools still don't appear after restart: 1. **Check Cursor logs**: Help > Toggle Developer Tools > Console - Look for MCP server errors - Check if `code-intelligence` server is listed 2. **Verify server starts manually**: ```powershell C:\AIStack-MCP\.venv\Scripts\python.exe C:\AIStack-MCP\mcp_intelligence_server.py --workspace C:\AIStack ``` Should show: "Ready for MCP connections" 3. **Check workspace path**: The `${workspaceFolder}` variable should resolve to your current project path 4. **Verify dependencies**: Ensure Ollama and Qdrant are running

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/mjdevaccount/AIStack-MCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server