ragieai/mcp-server

Overview

Ragie Model Context Protocol Server is an open-source JavaScript server that implements the Model Context Protocol (MCP) for connecting AI agents to retrieval-augmented generation (RAG) pipelines. It exposes a standardized interface for agents to query and retrieve context from external knowledge bases via the Ragie API.

Best for

Best for
Developers building MCP-compatible AI agents that need RAG context from Ragie knowledge bases

Use cases

• Connect an MCP-compatible AI agent to a Ragie knowledge base for context retrieval
• Build a retrieval-augmented generation pipeline with a standardized protocol interface
• Integrate external document search into agent workflows without custom API code

How to use

Install

RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server

Tools exposed

• retrieve

Tested with

Cursor, Claude Desktop

Example client config

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

Notes

Ragie Model Context Protocol Server is an open-source JavaScript server that implements the Model Context Protocol (MCP) for connecting AI agents to retrieval-augmented generation (RAG) pipelines. It exposes a standardized interface for agents to query and retrieve context from external knowledge bases via the Ragie API.

86 stars on GitHub. Last updated 2026-02-04. Licensed MIT.

Use cases

• Connect an MCP-compatible AI agent to a Ragie knowledge base for context retrieval
• Build a retrieval-augmented generation pipeline with a standardized protocol interface
• Integrate external document search into agent workflows without custom API code

Pros

• Implements the emerging MCP standard for agent-to-tool communication
• Lightweight JavaScript server with minimal dependencies
• Open source with 86 stars and active community potential

Cons

• Requires a Ragie API key and account, adding a third-party dependency
• Limited to MCP-compatible agents, not a general-purpose RAG solution
• Small community and early-stage project with limited documentation

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.