Wikimedia Enterprise MCP Server

Knowledge & MemoryPython

Access Wikimedia content and data through the Wikimedia Enterprise API

Available Tools

search_wikipedia

Search for Wikipedia articles based on a query

querylimit

get_wikipedia_article

Retrieve the content of a specific Wikipedia article

title

get_wikidata_entity

Retrieve information about a Wikidata entity

entity_id

Wikimedia Enterprise MCP provides seamless access to Wikipedia and Wikidata content through the Wikimedia Enterprise API. This integration allows AI models to retrieve up-to-date information from Wikipedia articles and Wikidata entities, enhancing their knowledge base with reliable, structured data from the world's largest free knowledge repository.

Overview

The Wikimedia Enterprise MCP server provides AI models with access to Wikipedia articles and Wikidata entities through the Wikimedia Enterprise API. This allows models to retrieve current, accurate information directly from Wikimedia's knowledge repositories.

Prerequisites

Before installing the Wikimedia Enterprise MCP, you'll need:

A Wikimedia Enterprise account (free credentials available at https://enterprise.wikimedia.com/)
Python and Poetry installed on your system
Anthropic Desktop (if using with Claude)

Installation Options

Option 1: Using uvx (Recommended)

The simplest way to install and use the Wikimedia Enterprise MCP is through the uvx package manager:

uvx wikimedia-enterprise-mcp@latest --username "YOUR_USERNAME" --password "YOUR_PASSWORD"

Option 2: Manual Installation

To install from source:

Clone the repository:

git clone https://gitlab.wikimedia.org/repos/future-audiences/wikimedia-enterprise-model-context-protocol.git
cd wikimedia-enterprise-model-context-protocol

Create a .env file in the project directory with your credentials:
```
WME_USERNAME=your_username
WME_PASSWORD=your_password
```
Install dependencies using Poetry:
```
poetry install
```
Run the server:
```
poetry run start.sh
```

Configuration

Claude Desktop Configuration

To use this MCP with Claude Desktop, add it to your claude_desktop_config.json file:

{
  "mcpServers": {
    "wikimedia-enterprise": {
      "command": "uvx",
      "args": [
        "wikimedia-enterprise-mcp@latest",
        "--username",
        "YOUR_USERNAME",
        "--password",
        "YOUR_PASSWORD"
      ]
    }
  }
}

Replace YOUR_USERNAME and YOUR_PASSWORD with your Wikimedia Enterprise credentials.

Usage

Once configured, the MCP server allows AI models to:

Retrieve Wikipedia article content
Access Wikidata entity information
Get structured data about specific topics

The MCP handles authentication and API interactions automatically, providing the model with clean, formatted data from Wikimedia sources.

Troubleshooting

If you encounter issues:

Verify your Wikimedia Enterprise credentials are correct
Ensure you have the latest version of the package
Check that your environment has the necessary permissions to make API calls
For authentication issues, try logging in manually to the Wikimedia Enterprise portal to confirm your account status