SAKH MCP Server (uk.sadiqoon/sakh)

The SAKH server is a remote Model Context Protocol service that provides researchers direct access to the corpus of Sayyid Ali Khamenei. It exposes juridical rulings (fatwas), lectures, and intellectual writings in Arabic and Persian via seven read-only tools.

🛠️ Key Features

Uses the Model Context Protocol
Access to a corpus containing fatwas, lectures, and intellectual writings
Content language: Arabic/Persian
Provides seven read-only tools

🚀 Use Cases

Research access to Ayatollah Khamenei materials
Studying juridical rulings and lectures
Working with a knowledge graph as part of the corpus

⚡ Developer Benefits

Integrates through Streamable HTTP transport
Connection endpoint documented for MCP usage
Read-only tool interface suitable for research retrieval

⚠️ Limitations

Read-only access only (no write/update tools)
Repo documentation notes it does not include the server’s source code

🌐 Connection Details

Server URL: https://mcp.khamenei-online.org/mcp/sakh
Transport: Streamable HTTP

SAKH — Ayatollah Khamenei Research Corpus (MCP Server)

A remote Model Context Protocol server that gives researchers direct access to the corpus of Sayyid Ali Khamenei — juridical rulings, lectures and intellectual writings in Arabic and Persian — through seven read-only tools.

This repository hosts the public listing and documentation for the SAKH MCP server. The server itself is hosted and operated by SADIQOON TECHNOLOGIES LTD; this repo does not contain its source.

Connect

Server URL: https://mcp.khamenei-online.org/mcp/sakh
Transport: Streamable HTTP
Authentication: OAuth 2.1 with Dynamic Client Registration (PKCE S256)
Website & documentation: https://khamenei-online.org/mcp/
Official MCP Registry: uk.sadiqoon/sakh

Access is granted on request via Google sign-in. Register at https://n8n.sadiqoon.com/form/SAKH-MCP-request and your access is enabled immediately.

What it does

Retrieval is hybrid: dense multilingual embeddings combined with sparse keyword matching, fused by reciprocal rank and then re-ranked, so a question phrased in ordinary language reaches the right passage even when the wording differs from the source.

Alongside the text corpus sits a knowledge graph of roughly 145,000 concepts joined by evidence-bearing relations (grounded-in, opposes, critiques, derives-from), which lets a researcher trace the structure of an argument rather than only collect quotations.

Every answer carries its source — book, page and chapter — and the system reports its own match confidence rather than presenting weak results as strong. It never composes a ruling that is not in the corpus.

Tools

All seven are read-only; none is destructive.

Tool	Purpose
`sakh_about`	System card: methodology, evidence tiers, and when to use each tool
`sakh_get_fatwa`	Retrieves a juridical ruling verbatim with its source and match confidence
`sakh_search_corpus`	Hybrid semantic search across lectures, writings and jurisprudence
`sakh_graph_explore`	Traverses documented relations between concepts
`sakh_get_communities`	Panoramic survey of the thematic clusters in a domain
`sakh_subject_analysis`	Clarifies an external subject before a ruling is applied to it
`sakh_glossary_lookup`	Explains a juridical term from a composite glossary

Example prompts

What is Ayatollah Khamenei's ruling on a traveller who intends to stay ten days in one place?
What has Ayatollah Khamenei said about the resistance economy? Cite the sources.
Map the conceptual field around wilayat al-faqih: what does it rest on, and what does it stand against?

Publisher

SADIQOON TECHNOLOGIES LTD — registered in England and Wales, company number 17120720, 167–169 Great Portland Street, London, W1W 5PF. Contact: info@sadiqoon.uk

Intended use

Built for university faculty and researchers in Islamic studies and the humanities. Read-only access to a closed, self-hosted corpus.

License

Documentation in this repository is released under the MIT License. The corpus and the hosted service are proprietary to SADIQOON TECHNOLOGIES LTD.

SAKH — Ayatollah Khamenei Research Corpus