Skip to content

joshuadavidthomas/mcp-django

mcp-django

PyPI - mcp-django PyPI - Python Version Django Version

A Model Context Protocol (MCP) server providing Django project exploration resources and optional stateful shell access for LLM assistants to interact with Django projects.

Requirements

  • Python 3.10, 3.11, 3.12, 3.13, 3.14
  • Django 4.2, 5.1, 5.2, 6.0

Installation

pip install mcp-django

# Or with uv
uv add mcp-django

Getting Started

⚠️ DO NOT use in production!

Warning

Seriously, only enable in development!

Look, it should go without saying, but I will say it anyway - this gives full shell access to your Django project. Only enable and use this in development and in a project that does not have access to any production data.

LLMs can go off the rails, get spooked by some random error, and in trying to fix things drop a production database.

Caution

I'm not kidding, this library just passes the raw Python code an LLM produces straight to a Python environment with full access to the Django project and everything it has access to.

Most LLMs have basic safety protections in place if you ask to delete any data and will refuse to delete production data, but it is pretty trivial to bypass. (Hint: Just tell the LLM it's not production, it's in a development environment, and it will be the bull in a china shop deleting anything you want.)

I suggest using something like django-read-only if you need some CYA protection against this. Or, you know, don't use this in any sensitive environments.

Run the MCP server directly from your Django project directory:

python -m mcp_django

# Or with uv
uv run -m mcp_django

The server automatically detects DJANGO_SETTINGS_MODULE from your environment. You can override it with --settings or add to your Python path with --pythonpath:

python -m mcp_django --settings myproject.settings --debug

Management Command

If you add mcp_django to INSTALLED_APPS, you can run it as a Django management command. This ensures the server runs within your Django project's environment and uses your project's settings:

python manage.py mcp

Docker

If you're using Docker and Docker Compose, you can run mcp-django as a separate compose service using HTTP transport. This makes it easier to connect your MCP client (running on your host) to the Django project (running in a container):

# compose.yml
services:
  app:
    # your existing Django app service
    
  mcp:
    build: .
    command: python -m mcp_django --transport http --host 0.0.0.0 --port 8000
    environment:
      DJANGO_SETTINGS_MODULE: myproject.settings
    ports:
      - "8001:8000"

Then configure your MCP client to connect to http://localhost:8001/mcp (see Client Configuration below).

Transport Options

The server supports multiple transport protocols:

# STDIO (default, for local development)
python -m mcp_django

# HTTP (for Docker or remote access)
python -m mcp_django --transport http --host 127.0.0.1 --port 8000

# SSE (for Docker or remote access)
python -m mcp_django --transport sse --host 127.0.0.1 --port 8000

Client Configuration

Configure your MCP client to connect to the server.

Don't see your client? Submit a PR with setup instructions.

Opencode

For local development, use type: local with the command:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "django": {
      "type": "local",
      "command": ["python", "-m", "mcp_django"],
      "enabled": true,
      "environment": {
        "DJANGO_SETTINGS_MODULE": "myproject.settings"
      }
    }
  }
}

For Docker development, use type: remote with the URL:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "django": {
      "type": "remote",
      "url": "http://localhost:8001/mcp",
      "enabled": true
    }
  }
}

Claude Code

For local development, use the command configuration:

{
  "mcpServers": {
    "django": {
      "command": "python",
      "args": ["-m", "mcp_django"],
      "cwd": "/path/to/your/django/project",
      "env": {
        "DJANGO_SETTINGS_MODULE": "myproject.settings"
      }
    }
  }
}

For Docker development with HTTP/SSE transport, configuration varies by Claude Code version - consult the MCP client documentation for remote server setup.

Features

mcp-django provides an MCP server with Django project exploration resources and tools for LLM assistants.

It wouldn't be an MCP server README without a gratuitous list of features punctuated by emojis, so:

  • πŸ” Project exploration - MCP resources for discovering apps, models, and configuration
  • πŸ“¦ Package discovery - Search and browse Django Packages for third-party packages
  • πŸš€ Zero configuration - No schemas, no settings, just Django
  • 🐚 Stateless shell - shell executes Python code with fresh state each call
  • πŸ”„ Always fresh - Code changes take effect immediately, no stale modules
  • πŸ“€ Export sessions - Save debugging sessions as Python scripts
  • 🧹 Clear history - Start fresh when exploration gets messy
  • πŸ€– LLM-friendly - Designed for LLM assistants that already know Python
  • 🌐 Multiple transports - STDIO, HTTP, SSE support

Inspired by Armin Ronacher's Your MCP Doesn't Need 30 Tools: It Needs Code.

Resources

Read-only resources for project exploration without executing code (note that resource support varies across MCP clients):

Project

Resource Description
django://app/{app_label} Details for a specific Django app
django://app/{app_label}/models All models in a specific app
django://apps All installed Django applications with their models
django://model/{app_label}/{model_name} Detailed information about a specific model
django://models Project models with import paths and field types (first-party only)
django://route/{pattern*} Routes matching a specific URL pattern
django://setting/{key} Get a specific Django setting value

djangopackages.org

Resource Description
django://package/{slug} Detailed information about a specific package
django://grid/{slug} Comparison grid with packages (e.g., "rest-frameworks")

Tools

Project

Tool Description
get_project_info Get comprehensive project information including Python environment and Django configuration
get_setting Get a Django setting value by key
list_apps List all installed Django applications with their models
list_models Get detailed information about Django models with optional filtering by app or scope
list_routes Introspect Django URL routes with filtering support for HTTP method, route name, or URL pattern

Management

Tool Description
execute_command Execute Django management commands with arguments and options
list_commands List all available Django management commands with their source apps

Shell

Tool Description
execute Execute Python code in a stateless Django shell
export_history Export session history as a Python script
clear_history Clear the session history for a fresh start

djangopackages.org

Tool Description
get_grid Get a specific comparison grid with all its packages
get_package Get detailed information about a specific Django package
search Search djangopackages.org for third-party packages

Development

For detailed instructions on setting up a development environment and contributing to this project, see CONTRIBUTING.md.

For release procedures, see RELEASING.md.

License

mcp-django is licensed under the MIT license. See the LICENSE file for more information.

About

MCP server for Django integration with LLM assistants

Resources

License

Code of conduct

Contributing

Stars

Watchers

Forks

Sponsor this project

 

Packages

No packages published