RAGBrain

Provides programmatic access to RAGBrain namespaces, semantic search, and document retrieval via an MCP interface for Claude Desktop.

python

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": {
    "ragbrain-ragbrain-mcp": {
      "command": "ragbrain-mcp",
      "args": [],
      "env": {
        "RAGBRAIN_URL": "http://localhost:8000",
        "RAGBRAIN_TIMEOUT": "60",
        "RAGBRAIN_LOG_LEVEL": "INFO",
        "RAGBRAIN_MAX_RESULTS": "20",
        "RAGBRAIN_MAX_DOCUMENT_LENGTH": "100000"
      }
    }
  }
}

You set up the RAGBrain MCP to bridge Claude Desktop with your RAGBrain knowledge base, enabling semantic search and document access directly from your chat prompts. This MCP runs locally and exposes a simple, standardized interface so Claude Desktop can send queries and retrieve results from RAGBrain.

How to use

Install the MCP and configure Claude Desktop to load it as a local MCP server. Once configured, you can ask Claude Desktop to list namespaces, search for topics, browse documents in a namespace, retrieve documents by ID, or discover documents by topic. Your MCP URL (http://localhost:8000 by default) is used by the MCP to reach the RAGBrain API.

How to install

Prerequisites: Python is installed on your system.

Install the RAGBrain MCP package.

pip install ragbrain-mcp

Configuration and usage notes

Configure Claude Desktop to load the RAGBrain MCP by adding an entry under mcpServers in your Claude config. The example below assumes you will run the MCP locally and that the RAGBrain API is reachable at http://localhost:8000.

{
  "mcpServers": {
    "ragbrain": {
      "command": "ragbrain-mcp",
      "env": {
        "RAGBRAIN_URL": "http://localhost:8000"
      }
    }
  }
}

Tools available through this MCP

You can interact with RAGBrain through these endpoints provided by the MCP: list namespaces, perform semantic searches, browse docs in a namespace, retrieve a document by ID, and discover documents by topic.

Troubleshooting

If Claude Desktop cannot connect, ensure the RAGBrain service is running and reachable at the configured URL. For example, verify http://localhost:8000/health responds correctly.

If tools aren’t appearing, confirm the MCP is correctly loaded by Claude Desktop and that the mcpServers entry is present and correctly named as ragbrain.

Available tools

ragbrain_list_namespaces

List namespaces with document counts to understand the scope of your knowledge base.

ragbrain_search

Perform semantic search across documents to find relevant information quickly.

ragbrain_browse_namespace

List all documents within a specific namespace to explore available content.

ragbrain_get_document

Retrieve the full content of a document by its ID for in-depth reading.

ragbrain_discover_documents

Discover documents by topic using summary search to surface related content.

Built by

VeilStrat

AI signals for GTM teams