Sefaria Jewish Library MCP Server
@Sivan22
Fetch and read Jewish texts through the API of Sefaria.org
Overview
What is Sefaria Jewish Library MCP Server?
The Sefaria Jewish Library MCP Server is a Model Context Protocol (MCP) server that provides access to Jewish texts from the Sefaria library. It enables Large Language Models to retrieve and reference Jewish texts through a standardized interface, intended for developers and users who want to integrate Jewish textual resources into AI-driven applications.
How to use Sefaria Jewish Library MCP Server?
Requires Python 3.10 or higher. Clone the repository, then run with uv --directory path/to/directory run sefaria_jewish_library. For Claude Desktop or Cline, add the server configuration as shown in the README. Can also be installed automatically via Smithery using npx -y @smithery/cli install mcp-sefaria-server --client claude. Once running, the server exposes four MCP tools.
Key features of Sefaria Jewish Library MCP Server
- Retrieve Jewish texts by reference (e.g., "Genesis 1:1")
- Retrieve commentaries on a given text
- Search the Jewish library for a query
- Get daily/weekly learning schedule from Sefaria's calendar
Use cases of Sefaria Jewish Library MCP Server
- Assist users in studying Torah with direct text lookup
- Provide commentaries alongside biblical or rabbinic passages
- Enable AI to answer questions using authentic Jewish sources
- Deliver daily learning schedules (e.g., Daf Yomi, Parashat Hashavua)
FAQ from Sefaria Jewish Library MCP Server
What are the runtime requirements?
Python 3.10 or higher is required. The project uses the MCP SDK (>=1.1.1) and the Sefaria API.
How do I install the server?
Clone the repository, then run via uv or configure it in an MCP client (e.g., Claude Desktop) using the provided JSON configuration. You can also use Smithery for automatic installation.
Where does the text data come from?
All texts are retrieved live from the Sefaria API; no local database is used. The server queries Sefaria’s public API endpoints.
What tools are available?
There are four tools: get_text (retrieve a text by reference), get_commentaries (get commentaries for a given reference), search_texts (search the library by query), and get_daily_learnings (return the daily or weekly learning schedule).
Does the server support authentication or custom configurations?
No authentication is required. For the get_daily_learnings tool, you can optionally specify parameters like diaspora, custom, year, month, day, and timezone to customize the schedule returned.
