ryge/wikidata-mcp

Wikidata SPARQL MCP Server

A Model Context Protocol (MCP) server that provides access to Wikidata's SPARQL endpoint, enabling AI assistants to query the world's largest free knowledge base for facts and structured information.

Features

• 🔍 SPARQL Query Execution: Execute custom SPARQL queries against Wikidata's knowledge base
• 🔎 Entity Search: Search for Wikidata entities by name across multiple languages
• 📊 Entity Information: Get detailed information about specific Wikidata entities
• 📚 Example Queries: Built-in collection of useful SPARQL query examples
• 🌐 Dual Transport Support: Works with both stdio and HTTP (SSE) transports
• ⚡ Async Support: Built with modern async/await patterns for optimal performance

What is Wikidata?

Wikidata is a free, collaborative knowledge base that can be read and edited by both humans and machines. It contains over 100 million items with structured data about people, places, concepts, and more. The SPARQL endpoint allows you to query this vast knowledge base using semantic queries.

Installation

Prerequisites

• Python 3.10 or higher
• pip or uv package manager

Install Dependencies

Using pip:

pip install -r requirements.txt

Using uv (recommended):

uv pip install -r requirements.txt

Usage

Running with stdio (default)

For use with Claude Desktop or other MCP clients:

python server.py

Running with Python virtual environments

If you're using a Python virtual environment (recommended), you can use the provided shell script that automatically activates the virtual environment, runs the server, and deactivates:

./server.sh

This script assumes your virtual environment is located at .venv/ in the project directory. If you use a different location, you'll need to modify server.sh accordingly.

Running with HTTP transport

For web-based access or testing:

python server.py --http

The server will start on http://localhost:8000

Configuration for Claude Desktop

Add the following to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "wikidata": {
      "command": "python",
      "args": [
        "/Users/anr/Documents/Claude/wikidata-mcp/server.py"
      ]
    }
  }
}

Or using uv:

{
  "mcpServers": {
    "wikidata": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "fastmcp",
        "--with",
        "httpx",
        "/Users/anr/Documents/Claude/wikidata-mcp/server.py"
      ]
    }
  }
}

Available Tools

1. sparql_query

Execute SPARQL queries against Wikidata.

Parameters:

• query (string, required): The SPARQL query to execute
• timeout (int, optional): Query timeout in seconds (default: 30, max: 60)
• format (string, optional): Response format - 'json' or 'xml' (default: 'json')

Example:

await sparql_query("""
  SELECT ?person ?personLabel WHERE {
    ?person wdt:P31 wd:Q5 .
    ?person wdt:P19 wd:Q78 .
    SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
  }
  LIMIT 10
""")

2. search_entities

Search for Wikidata entities by name.

Parameters:

• search_term (string, required): The term to search for
• language (string, optional): Language code (default: 'en')
• limit (int, optional): Max results (default: 10, max: 50)
• type (string, optional): Filter by 'item' or 'property'

Example:

await search_entities("Albert Einstein", language="en", limit=5)

3. get_entity_info

Get detailed information about a specific Wikidata entity.

Parameters:

• entity_id (string, required): The Wikidata ID (e.g., 'Q937')
• language (string, optional): Language code (default: 'en')

Example:

await get_entity_info("Q937", language="en")  # Albert Einstein

Available Resources

1. wikidata://examples/queries

Get example SPARQL queries for common use cases.

2. wikidata://help/prefixes

Get documentation about common SPARQL prefixes used in Wikidata.

Example Queries

Find Nobel Prize Winners

SELECT ?person ?personLabel ?awardYear WHERE {
  ?person wdt:P166 wd:Q38104 .
  OPTIONAL { ?person p:P166 [ps:P166 wd:Q38104; pq:P585 ?awardYear] . }
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
}
ORDER BY DESC(?awardYear)
LIMIT 50

Get Properties of an Entity

SELECT ?property ?propertyLabel ?value ?valueLabel WHERE {
  wd:Q937 ?p ?value .
  ?property wikibase:directClaim ?p .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
}
LIMIT 50

Find Countries and Capitals

SELECT ?country ?countryLabel ?capital ?capitalLabel WHERE {
  ?country wdt:P31 wd:Q3624078 .
  ?country wdt:P36 ?capital .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
}
ORDER BY ?countryLabel

Common Wikidata Properties

• P31: instance of
• P279: subclass of
• P19: place of birth
• P20: place of death
• P569: date of birth
• P570: date of death
• P166: award received
• P106: occupation
• P27: country of citizenship
• P36: capital

Common Wikidata Entities

• Q5: human
• Q515: city
• Q6256: country
• Q3624078: sovereign state
• Q38104: Nobel Prize in Physics

Troubleshooting

Query Timeout

If queries timeout, try:

• Adding a LIMIT clause to reduce results
• Simplifying complex queries
• Increasing the timeout parameter (max 60 seconds)

Empty Results

• Verify entity IDs are correct (e.g., Q937, not 937)
• Check that properties exist for the entity
• Ensure SPARQL syntax is correct
• Add the SERVICE wikibase:label clause for human-readable labels

Connection Issues

• Verify internet connectivity
• Check if query.wikidata.org is accessible
• Ensure firewall isn't blocking requests

Testing

You can test the server using the MCP Inspector:

# Install MCP Inspector
npm install -g @modelcontextprotocol/inspector

# Start the server
python server.py

# In another terminal, start the inspector
mcp-inspector

Resources

• Wikidata Query Service
• SPARQL Tutorial
• Wikidata Property Browser
• FastMCP Documentation
• Model Context Protocol

License

This project is licensed under the BSD 2-Clause License - see the file for details.

This MCP server is provided for use with Wikidata's public SPARQL endpoint. Please respect Wikidata's query limits and terms of service.

Contributing

Feel free to enhance this server with additional tools and features:

• Support for federated queries
• Query result caching
• Query optimization suggestions
• Additional entity relationship tools
• Batch query support

Support

For issues with:

• This MCP server: Check the server logs and error messages
• Wikidata queries: Consult Wikidata SPARQL documentation
• MCP protocol: See MCP documentation