Genealogy research with AI assistants is limited and frustrating:
- No direct access to your family tree data
- Manual data entry and research across multiple platforms
- Generic genealogy advice without context of your specific family
- No ability to automatically update or maintain your research
Gramps MCP provides AI assistants with direct access to your Gramps genealogy database through a comprehensive set of tools. Your AI assistant can now:
- Smart Search: Find people, families, events, places, and sources across your entire database
- Data Management: Create and update genealogy records with proper validation
- Tree Analysis: Trace descendants, ancestors, and family connections
- Relationship Discovery: Explore family connections and research gaps
- Tree Information: Get comprehensive tree statistics and track changes
Add Gramps MCP to your AI assistant and transform how you research family history:
Search for all descendants of John Smith born in Ireland before 1850Create a new person record for Mary O'Connor with birth date 1823 in County CorkFind all families missing marriage dates and suggest research prioritiesNo more manual data entry, no context switching between apps, no generic genealogy advice.
- Connect to your Gramps Web API
- Install Gramps MCP in your AI assistant
- Start intelligent genealogy research with natural language
- find_type - Universal search for any entity type (person, family, event, place, source, citation, media, repository) using Gramps Query Language
- find_anything - Text search across all genealogy data (matches literal text, not logical combinations)
- get_type - Get comprehensive information about specific persons or families by ID
- create_person - Create or update person records
- create_family - Create or update family units
- create_event - Create or update life events
- create_place - Create or update geographic locations
- create_source - Create or update source documents
- create_citation - Create or update citations
- create_note - Create or update textual notes
- create_media - Create or update media files
- create_repository - Create or update repository records
- tree_stats - Get tree statistics and information
- get_descendants - Find all descendants of a person
- get_ancestors - Find all ancestors of a person
- recent_changes - Track recent modifications to your data
- Gramps Web server with your family tree data - Setup Guide
- Docker and Docker Compose
- MCP-compatible AI assistant (Claude Desktop, Cursor, etc.)
-
Ensure Gramps Web is Running:
- Follow the Gramps Web setup guide to get your family tree online
- Note your Gramps Web URL, username, and password
- Find your tree ID under System Information in your Gramps Web interface
-
Start the Server:
# Download the configuration
curl -O https://raw.githubusercontent.com/cabout-me/gramps-mcp/main/docker-compose.yml
curl -O https://raw.githubusercontent.com/cabout-me/gramps-mcp/main/.env.example
cp .env.example .env
# Edit .env with your Gramps Web API credentials
# Start the server
docker-compose up -dThat's it! The MCP server will be running at http://localhost:8000/mcp
If you prefer to run the server directly with Python:
- Setup Python Environment:
# Install uv (if not already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh
# Install dependencies
uv sync- Run the Server:
# HTTP transport (for web-based MCP clients)
uv run python -m src.gramps_mcp.server
# Stdio transport (for CLI-based MCP clients)
uv run python -m src.gramps_mcp.server stdioThe HTTP server will be available at http://localhost:8000/mcp, while stdio runs directly in the terminal.
Create a .env file with your Gramps Web settings:
# Your Gramps Web instance (from step 1)
GRAMPS_API_URL=https://your-gramps-web-domain.com # Without /api suffix - will be added automatically
GRAMPS_USERNAME=your-gramps-web-username
GRAMPS_PASSWORD=your-gramps-web-password
GRAMPS_TREE_ID=your-tree-id # Find this under System Information in Gramps WebAdd to your Claude Desktop MCP configuration file (claude_desktop_config.json):
Using Docker (works with both pre-built and local images):
{
"mcpServers": {
"gramps": {
"command": "docker",
"args": ["exec", "-i", "gramps-mcp-gramps-mcp-1", "python", "-m", "src.gramps_mcp.server", "stdio"]
}
}
}Using uv directly (if running without Docker):
{
"mcpServers": {
"gramps": {
"command": "uv",
"args": ["run", "python", "-m", "src.gramps_mcp.server", "stdio"],
"cwd": "/path/to/gramps-mcp"
}
}
}OpenWebUI recommends using the mcpo proxy to expose MCP servers as OpenAPI endpoints.
With uv:
uvx mcpo --port 8000 -- uv run python -m src.gramps_mcp.server stdioWith Docker:
uvx mcpo --port 8000 -- docker exec -i gramps-mcp-gramps-mcp-1 uv run python -m src.gramps_mcp.server stdioHTTP Transport:
claude mcp add --transport http gramps http://localhost:8000/mcpStdio Transport (direct connection, more efficient):
# Using Docker
claude mcp add --transport stdio gramps "docker exec -i gramps-mcp-gramps-mcp-1 sh -c 'cd /app && python -m src.gramps_mcp.server stdio'"
# Using uv directly (requires local setup)
claude mcp add --transport stdio gramps "uv run python -m src.gramps_mcp.server stdio"Transport Choice: Use stdio for better performance and direct integration with CLI tools like Claude Code. Use HTTP when you need the server to handle multiple clients or prefer web-based access.
For any other MCP client, use the HTTP transport endpoint:
{
"mcpServers": {
"gramps": {
"url": "http://localhost:8000/mcp"
}
}
}src/gramps_mcp/
|-- server.py # MCP server with HTTP transport
|-- tools.py # Tool registry and exports
|-- client.py # Gramps Web API client
|-- models.py # Pydantic data models
|-- auth.py # JWT authentication
|-- config.py # Configuration management
|-- tools/ # Modular tool implementations
| |-- search_basic.py
| |-- search_details.py
| |-- data_management.py
| |-- tree_management.py
| `-- analysis.py
|-- handlers/ # Data formatting handlers
`-- client/ # API client modules
- MCP Python SDK: Model Context Protocol implementation
- FastAPI: HTTP server for MCP transport
- Pydantic: Data validation and serialization
- httpx: Async HTTP client for API communication
- PyJWT: JWT token authentication
- python-dotenv: Environment configuration
Find all people with the surname "Smith" born in IrelandShow me recent changes to the family tree in the last 30 daysCreate a new person record for Patrick O'Brien, born 1845 in Cork, IrelandAdd a marriage event for John and Mary Smith on June 15, 1870 in BostonFind all descendants of Margaret Kelly and show their birth locationsShow me statistics about my family tree - how many people, families, and eventsWhat recent changes have been made to my family tree in the last week?- JWT token authentication with automatic refresh
- Environment-based credential management
- Input validation using Pydantic models
- Secure HTTP transport with proper error handling
- No sensitive data exposed in tool responses
Connection refused errors: Ensure your Gramps Web API server is running and accessible at the configured URL.
Authentication failures: Verify your username and password are correct and the user has appropriate permissions.
Tool timeout errors: Check your network connection and consider increasing timeout values for large datasets.
Docker issues: Ensure Docker and Docker Compose are installed and running.
To enable debug logging, check your application logs with:
docker-compose logs -fThis project is licensed under the GNU Affero General Public License v3.0 - see the LICENSE file for details.
- Gramps - Free genealogy software
- Gramps Web API - Web API for Gramps
- Model Context Protocol - Standard for AI tool integration
We welcome contributions! Please see our Contributing Guide for details on:
- Setting up the development environment
- Running tests and maintaining code quality
- Submitting pull requests
- Reporting issues and requesting features
- Bug Reports & Feature Requests: GitHub Issues
- Questions & Discussions: GitHub Discussions
- Documentation: Project Wiki
- The Gramps Project team for creating excellent genealogy software
- Anthropic for developing the Model Context Protocol
- The genealogy research community for inspiration and feedback