DRF MCP Docs

Install the package via pip. Use the `[spectacular]` extra if your project uses drf-spectacular (recommended), or `[yasg]` for drf-yasg. The base install works with DRF's built-in schema generation.

pip install drf-mcp-docs

# With drf-spectacular (recommended)
pip install drf-mcp-docs[spectacular]

# With drf-yasg
pip install drf-mcp-docs[yasg]

Add drf_mcp_docs to your Django INSTALLED_APPS. No additional settings are required for basic use — the package auto-detects your configured schema generator.

INSTALLED_APPS = [
    # ...
    'rest_framework',
    'drf_mcp_docs',
]

Run the checkmcpconfig management command to validate that the schema generator is detected and documentation can be generated successfully before connecting any AI client.

python manage.py checkmcpconfig

Run the MCP server using stdio transport for local AI tools (Claude Code, Cursor) or streamable-HTTP for remote network access.

# stdio transport — for local AI tools
python manage.py runmcpserver --transport stdio

# Streamable HTTP transport — for network access
python manage.py runmcpserver --transport streamable-http --host 0.0.0.0 --port 8100

Add the server to your MCP client configuration. For Claude Code, add to ~/.claude.json. For Cursor, add to .cursor/mcp.json. Use the absolute path to your Django project.

{
  "mcpServers": {
    "my-api-docs": {
      "command": "python",
      "args": ["manage.py", "runmcpserver", "--transport", "stdio"],
      "cwd": "/path/to/your/django/project"
    }
  }
}