Noor MCP Code Assistant
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Noor MCP Code AssistantHow is pagination implemented in my project?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
🚀 Noor MCP Code Assistant
MCP (Model Context Protocol) Server that bridges Claude Desktop with your RAG-powered Code Agent API.
📋 Overview
This project creates an MCP server that allows Claude Desktop to communicate with your existing RAG Code Agent API running at localhost:8900. When you ask code-related questions in Claude Desktop, it can use this tool to fetch context-aware answers from your codebase.
🎯 Features
✅ Seamless integration with Claude Desktop
✅ Connects to existing RAG API (
/api/chat/rag)✅ Session management for conversation continuity
✅ Comprehensive error handling
✅ Configurable via environment variables
✅ Ready for future agent extensions (SQL, React, Java)
Related MCP server: Claude Code Connector MCP
🏗️ Project Structure
noor_mcp_code_assistant_parent/
├── env_container/ # Virtual environment
│ ├── bin/
│ ├── lib/
│ └── ...
└── noor_mcp_code_assistant/ # Project source
├── mcp_server.py # Main MCP server
├── config.py # Configuration management
├── requirements.txt # Dependencies
├── .env.example # Environment template
├── .env # Your configuration (git-ignored)
└── README.md # This file🚀 Quick Start
1️⃣ Create Virtual Environment
Navigate to the parent folder and create the environment:
cd noor_mcp_code_assistant_parent
python3.11 -m venv env_containerActivate the environment:
# Linux/WSL/macOS
source env_container/bin/activate
# Windows PowerShell
.\env_container\Scripts\Activate2️⃣ Install Requirements
cd noor_mcp_code_assistant
pip install -r requirements.txt3️⃣ Configure Environment Variables
Copy the example file and customize:
cp .env.example .envEdit .env with your settings:
# Required
RAG_API_BASE_URL=http://localhost:8900
RAG_CHAT_ENDPOINT=/api/chat/rag
# Optional
DEFAULT_SESSION_ID=claude-desktop-session
REQUEST_TIMEOUT=1204️⃣ Configure Claude Desktop
Locate your Claude Desktop configuration file:
Windows:
%APPDATA%\Claude\claude_desktop_config.json(Usually: C:\Users\<YourUsername>\AppData\Roaming\Claude\claude_desktop_config.json)
Add the MCP server configuration:
{
"mcpServers": {
"noor-code-assistant": {
"command": "wsl",
"args": [
"-d", "Ubuntu-20.04",
"--",
"/home/<your-username>/noor_mcp_code_assistant_parent/env_container/bin/python",
"/home/<your-username>/noor_mcp_code_assistant_parent/noor_mcp_code_assistant/mcp_server.py"
],
"env": {
"PYTHONUNBUFFERED": "1"
}
}
}
}⚠️ Important: Replace
<your-username>with your actual WSL username and adjust paths accordingly.
🔍 Find Your WSL Distribution Name
wsl --list --verboseUse the exact name shown (e.g., Ubuntu-20.04, Ubuntu, etc.)
5️⃣ Verify Setup
Step 1: Ensure your RAG API is running:
# In WSL terminal
cd /path/to/your/rag-project
source env_container/bin/activate
uvicorn main:app --reload --port 8900Step 2: Test the RAG API:
curl -X POST http://localhost:8900/api/chat/rag \
-H "Content-Type: application/json" \
-d '{"session_id": "test", "message": "hello"}'Step 3: Restart Claude Desktop completely (check system tray)
Step 4: The noor-code-assistant tool should now appear in Claude Desktop!
💬 Usage Examples
Once configured, you can ask Claude Desktop:
"Use the code assistant to create a C# function that gets employee requests by employee ID"
"Ask the code assistant how pagination is implemented in my project"
"Use code_assistant to show me the DTO structure for Employee entity"
🔮 Adding Future Agents
When you're ready to add more agents (SQL, React, Java):
Step 1: Add endpoints to your RAG API
@app.post("/api/sql/agent")
async def sql_agent(request: SQLAgentRequest):
# Your SQL agent logic
passStep 2: Update .env
SQL_AGENT_ENDPOINT=/api/sql/agentStep 3: Uncomment tool definitions in mcp_server.py
Look for the commented sql_agent tool and handler, then uncomment them.
Step 4: Restart Claude Desktop
🐛 Troubleshooting
Issue | Solution |
MCP server not appearing | Check Claude Desktop logs at |
Connection refused | Ensure RAG API is running on |
WSL path issues | Use full Linux paths (e.g., |
Python not found | Use full path to Python in venv |
Timeout errors | Increase |
📋 View Logs
Claude Desktop Logs:
C:\Users\<YourUsername>\AppData\Roaming\Claude\logs\Test MCP Server Standalone:
cd noor_mcp_code_assistant
source ../env_container/bin/activate
python mcp_server.py📊 Configuration Reference
Variable | Default | Description |
|
| Base URL of your RAG API |
|
| Chat endpoint path |
|
| Default session for continuity |
|
| API timeout in seconds |
🔐 Security Notes
This MCP server only makes outbound HTTP requests to your local RAG API
No sensitive data is stored or logged
Session IDs are used only for conversation context
All communication happens over localhost
📄 License
MIT License - Feel free to use and modify!
👨💻 Author
Noor Uddin
🎥 YouTube: Noor Codelogics
💼 LinkedIn: Noor Uddin
🛒 CodeCanyon: 180+ Sales
🌟 Support
If you find this helpful, please:
⭐ Star the repository
📺 Subscribe to Noor Codelogics
🔗 Share with fellow developers
Happy Coding! 🚀
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
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/Nooruddin-dev/noor_mcp_code_assistant'
If you have feedback or need assistance with the MCP directory API, please join our Discord server