Librarian

Provides Wikipedia querying and automatic fact-checking capabilities for MCP clients.

python

1

GitHub Stars

python

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "mlziade-librarian": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/your/librarian",
        "run",
        "python",
        "librarian_stdio.py"
      ],
      "env": {
        "PYTHONPATH": "/path/to/your/librarian"
      }
    }
  }
}

Librarian is an MCP Server that lets a compatible client query Wikipedia to fetch facts, summaries, and page details, and it can be configured to automatically fact-check information during conversations. This makes it easy to pull reliable Wikipedia-backed information into your workflows with proactive verification when desired.

How to use

You connect an MCP client to the Librarian server and use its tools to search Wikipedia, retrieve page information, summarize content, or extract specific sections. You can also enable automatic fact-checking so factual claims are proactively verified against Wikipedia without needing a direct prompt.

Key capabilities you can leverage include:

When you start a session, you can issue requests through your MCP client to call the available tools. The tools return top results, full page info, concise summaries, and section content that you can present to users or use as sources in your own workflows. If automatic fact-checking is enabled, verified corrections and source attributions are provided automatically.

How to install

Prerequisites you need before installing Librarian on your machine:

Python 3.13 or higher
uv package manager installed on your system
Claude Desktop installed if you plan to use it directly with Librarian for automatic fact-checking

Install and configure Librarian as an MCP server for Claude Desktop

{
  "mcpServers": {
    "librarian": {
      "command": "uv",
      "args": ["--directory", "/path/to/your/librarian", "run", "python", "librarian_stdio.py"],
      "env": {
        "PYTHONPATH": "/path/to/your/librarian"
      }
    }
  }
}

Note: Replace "/path/to/your/librarian" with the actual path where you placed the Librarian project. This configuration tells the MCP system to run Librarian using uv in the correct project environment.

Starting and testing the MCP server locally

Ensure Librarian is set up in your MCP configuration as shown above.
Start the MCP server configuration in your MCP client or environment. The command uses uv to launch the Librarian runtime with the Python entry point.
Verify the server is reachable by checking the MCP client logs for a successful connection and by issuing a simple Wikipedia search through the client.

Troubleshooting and tips

If you encounter issues, check the logs for any module or path errors, ensure the directory path in the config matches where Librarian is located, and confirm that the Python path is correctly set.

Common problems include module import errors or missing files. Use absolute paths in both the command and the environment configuration to reduce path-related failures.

Available tools

search_wikipedia_pages

Search for Wikipedia articles and return the top results with selection information.

get_wikipedia_page_info

Retrieve comprehensive information about a specific Wikipedia page, including content, summary, linked terms, and categories.

get_wikipedia_page_summary

Get a concise summary of a Wikipedia page with configurable length.

get_wikipedia_page_sections

List all sections from a Wikipedia page to access targeted information.

get_wikipedia_page_sections_info

Fetch detailed content for specific sections of a Wikipedia page by title or index.