Django SaaS MCP Server

An MCP (Model Context Protocol) server that provides autonomous assistance for Django SaaS development. This server runs continuously in the background, monitoring your Django project and providing proactive help through VS Code Copilot without requiring manual intervention.

Features

• 🔍 Continuous Monitoring: Automatically watches Django project files for changes
• 🤖 Autonomous Assistance: Provides suggestions and detects issues without user action
• 🛠️ Django-Specific Tools: Exposes specialized tools for Django development
• 📊 Project Analysis: Analyzes Django project structure and identifies apps
• ⚠️ Issue Detection: Detects common Django patterns, anti-patterns, and security issues
• 🔄 Real-time Updates: Monitors models, views, settings, templates, and more

What It Monitors

The server automatically monitors:

• Python files (**/*.py)
• Django core files (settings.py, models.py, views.py, urls.py, manage.py)
• Templates (**/templates/**/*.html)
• Requirements files (requirements*.txt)

And ignores:

• Virtual environments (venv/, env/)
• Python cache (__pycache__/, *.pyc)
• Node modules (node_modules/)
• Build directories

Available Tools

1. set_django_project_path

Set the path to your Django project for monitoring.

Input:

{
  "path": "/absolute/path/to/django/project"
}

Output:

{
  "success": true,
  "message": "Now monitoring Django project at: /path/to/project",
  "monitoringActive": true
}

2. analyze_django_project

Analyze the Django project structure and detect potential issues.

Output:

{
  "projectPath": "/path/to/project",
  "hasDjango": true,
  "apps": ["blog", "users", "api"],
  "hasSettings": true,
  "hasManagePy": true,
  "issues": [],
  "fileCount": 45
}

3. check_django_file

Check a specific Django file for common issues and get suggestions.

Input:

{
  "filePath": "myapp/models.py"
}

Output:

{
  "filePath": "/absolute/path/to/file",
  "exists": true,
  "issues": ["Missing Django models import"],
  "suggestions": ["Add: from django.db import models"]
}

4. list_monitored_files

List all files currently being monitored.

Input:

{
  "pattern": "models.py"  // Optional filter
}

Output:

{
  "totalFiles": 3,
  "files": [
    {
      "path": "/full/path/to/file",
      "relativePath": "blog/models.py",
      "size": 1024,
      "lastModified": "2025-10-26T12:00:00.000Z"
    }
  ]
}

Installation & Setup

Prerequisites

• Node.js 18+ installed
• VS Code with GitHub Copilot
• A Django project

Setup Steps

1. Install Dependencies (already done during workspace creation)

   npm install

2. Build the Server (already done)

   npm run build

3. Configure in VS Code The server is already configured in .vscode/mcp.json. VS Code Copilot will automatically connect to it when you open this workspace.

4. Verify Connection

   • Open VS Code
   • Open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P)
   • Type "MCP" to see MCP-related commands
   • The django-saas-assistant server should be listed

Usage

With VS Code Copilot

Once configured, the MCP server runs automatically in the background. You can interact with it through Copilot:

1. Set Your Django Project Path In Copilot chat, ask:

   @workspace Use the django-saas-assistant to monitor my Django project at /path/to/my/project

2. Analyze Your Project

   @workspace Analyze my Django project structure

3. Check for Issues

   @workspace Check my Django models file for issues

4. List Monitored Files

   @workspace Show me all monitored Django files

Autonomous Monitoring

The server runs continuously and logs activity to stderr (visible in the MCP output channel):

• File additions
• File modifications
• File deletions
• Detected issues

Development

Project Structure

.
├── src/
│   └── index.ts          # Main server implementation
├── build/                # Compiled JavaScript (generated)
├── .vscode/
│   └── mcp.json         # MCP server configuration
├── .github/
│   └── copilot-instructions.md
├── package.json
├── tsconfig.json
└── README.md

Building

npm run build

Watch Mode (for development)

npm run watch

Running Locally

npm run dev

Issue Detection

The server automatically detects:

Models (models.py)

• Missing Django models import
• Incorrect model class definitions
• Missing migrations directories

Views (views.py)

• Missing Django imports

Settings (settings.py)

• DEBUG = True with empty ALLOWED_HOSTS (security risk)
• Hardcoded SECRET_KEY (security risk)

Project Structure

• Settings file without manage.py
• Django project without apps

How It Works

1. File Watching: Uses chokidar to monitor Django files in real-time
2. Pattern Detection: Analyzes file content for Django-specific patterns
3. MCP Protocol: Exposes tools via the Model Context Protocol
4. Stdio Transport: Communicates with VS Code through standard input/output
5. Copilot Integration: Works seamlessly with GitHub Copilot for autonomous assistance

Debugging

To view server logs:

1. Open VS Code Output panel (View → Output)
2. Select "MCP: django-saas-assistant" from the dropdown
3. Watch real-time logs as the server monitors your project

Configuration

Changing the Project Path

Edit .vscode/mcp.json to point to a different location:

{
  "servers": {
    "django-saas-assistant": {
      "type": "stdio",
      "command": "node",
      "args": [
        "/your/custom/path/build/index.js"
      ]
    }
  }
}

Troubleshooting

Server Not Starting

• Ensure Node.js is installed: node --version
• Check build output: npm run build
• Verify the path in .vscode/mcp.json

No Files Being Monitored

• Use the set_django_project_path tool to set the correct path
• Check that the path contains Django files
• Review the MCP output logs

Tools Not Appearing in Copilot

• Restart VS Code
• Check MCP server status in the Output panel
• Verify .vscode/mcp.json configuration

Contributing

This is an autonomous assistant for Django SaaS development. To extend it:

1. Add new tools in src/index.ts using server.registerTool()
2. Implement new file watchers for additional patterns
3. Add more issue detection logic
4. Build and test: npm run build

License

MIT

Support

For issues or questions:

• Check the MCP server logs in VS Code Output
• Review this README
• Check the MCP specification: https://modelcontextprotocol.io/

Note: This server is designed to run continuously in the background. It automatically starts when you open VS Code with this workspace and provides autonomous assistance through Copilot without requiring manual commands.