📌Overview

Purpose: The Ragie Model Context Protocol (MCP) server provides a means for AI models to access and retrieve information from a Ragie knowledge base efficiently.

Overview: This server implements the Model Context Protocol, enabling users to query a customizable knowledge base with a single retrieval tool. It supports flexible command line options and easy setup for integration in various development environments.

Key Features:

Retrieve Tool: A powerful search function that allows users to query the knowledge base with customizable parameters (query string, topK, rerank, and recencyBias) to obtain relevant information.
Flexible Installation: Simple installation process using Node.js, with options to integrate the server across different platforms (e.g., Cursor, Claude desktop) through configuration files or shell scripts.

Ragie Model Context Protocol Server

A Model Context Protocol (MCP) server that provides access to Ragie's knowledge base retrieval capabilities.

Description

This server implements the Model Context Protocol to enable AI models to retrieve information from a Ragie knowledge base. It provides a single tool called retrieve that allows querying the knowledge base for relevant information.

Prerequisites

Node.js >= 18
A Ragie API key

Installation

Set the required environment variable:

RAGIE_API_KEY (required): Your Ragie API authentication key

Start the server on stdio for MCP protocol messages:

RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server

Command Line Options

--description, -d <text>: Override the default tool description with custom text
--partition, -p <id>: Specify the Ragie partition ID to query

Examples:

RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server --description "Search the company knowledge base for information"

RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server --partition your_partition_id

RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server --description "Search the company knowledge base" --partition your_partition_id

Cursor Configuration

To use this MCP server with Cursor:

Option 1: Create an MCP configuration file

Save a file called mcp.json

For tools specific to a project, create .cursor/mcp.json in your project directory.
For tools available across all projects, create ~/.cursor/mcp.json in your home directory.

Example mcp.json:

{
  "mcpServers": {
    "ragie": {
      "command": "npx",
      "args": [
        "-y",
        "@ragieai/mcp-server",
        "--partition",
        "optional_partition_id"
      ],
      "env": {
        "RAGIE_API_KEY": "your_api_key"
      }
    }
  }
}

Option 2: Use a shell script

Save a file ragie-mcp.sh:

#!/usr/bin/env bash

export RAGIE_API_KEY="your_api_key"

npx -y @ragieai/mcp-server --partition optional_partition_id

Make it executable: chmod +x ragie-mcp.sh
Add the MCP server script in Cursor under Settings → Cursor Settings → MCP Servers.

Replace your_api_key with your actual Ragie API key and set the partition ID if needed.

Claude Desktop Configuration

To use this MCP server with Claude desktop:

Create the MCP config file claude_desktop_config.json at:

MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json

Example config:

{
  "mcpServers": {
    "ragie": {
      "command": "npx",
      "args": [
        "-y",
        "@ragieai/mcp-server",
        "--partition",
        "optional_partition_id"
      ],
      "env": {
        "RAGIE_API_KEY": "your_api_key"
      }
    }
  }
}

Restart Claude desktop for changes to take effect.

The Ragie retrieval tool will then be available in your Claude desktop conversations.

Features

Retrieve Tool

The retrieve tool searches the knowledge base with parameters:

query (string): Search query
topK (number, optional, default: 8): Max number of results
rerank (boolean, optional, default: true): Return most relevant info
recencyBias (boolean, optional, default: false): Favor recent info

Returns an array of content chunks matching the query.

Development

This project is written in TypeScript using:

@modelcontextprotocol/sdk: MCP server implementation
ragie: Ragie API client
zod: Runtime type validation

Development setup

Run server in development mode:

RAGIE_API_KEY=your_api_key npm run dev -- --partition optional_partition_id

Build project:

npm run build

License

MIT License - See LICENSE.txt for details.

ragie-mcp-server