MCP.so
ログイン

Ragie MCP Server

@ragieai

基本情報

カテゴリ

その他

ライセンス

MIT

トランスポート

stdio

作者

ragieai

Ragie MCP Server について

Ragie Model Context Protocol Server

Ragie MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

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

Ragie MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンで一時的なサンドボックスでこのサーバーを実行し、ツール一覧を取得できます。

概要

What is Ragie Model Context Protocol Server?

The Ragie Model Context Protocol Server is an MCP server that provides AI models with access to Ragie’s knowledge base retrieval capabilities via a single “retrieve” tool. It is intended for developers integrating Ragie-powered knowledge retrieval into AI applications that support MCP.

How to use Ragie Model Context Protocol Server?

Install and run with npx using your Ragie API key: RAGIE_API_KEY=your_api_key npx @ragieai/mcp-server. Optionally, pass --description or --partition flags. Configure it in Cursor via an mcp.json file or a shell script, or in Claude Desktop via claude_desktop_config.json. The server listens on stdio for MCP protocol messages.

Key features of Ragie Model Context Protocol Server

  • Provides a single “retrieve” tool for knowledge base queries.
  • Accepts query, topK, rerank, and recencyBias parameters.
  • Returns an array of relevant content chunks.
  • Supports custom tool description and partition ID overrides.
  • Runs on Node.js >= 18 with a Ragie API key.
  • Works with Cursor and Claude Desktop MCP configurations.

Use cases of Ragie Model Context Protocol Server

  • Enable Claude Desktop to search a company knowledge base.
  • Let Cursor AI retrieve information from Ragie during development.
  • Integrate Ragie retrieval into any MCP-supporting AI assistant.
  • Query partitioned knowledge bases with the --partition option.

よくある質問

1

What does the “retrieve” tool do?

It searches the Ragie knowledge base for information matching a provided query and returns an array of text chunks, with optional reranking and recency bias.

2

What are the runtime requirements?

Node.js 18 or later and a valid Ragie API key (set as the `RAGIE_API_KEY` environment variable).

3

How do I configure it for Claude Desktop?

Create or edit `claude_desktop_config.json` (macOS: `~/Library/Application Support/Claude/…`, Windows: `%APPDATA%/Claude/…`) with the npx command, API key, and optional partition ID, then restart Claude.

4

Is this server still needed?

No. Ragie now supports MCP natively as a streamable HTTP server. This project is deprecated in favor of that built-in support.

5

Can I limit which part of the knowledge base is searched?

Yes. Use the `--partition` command-line option or set the partition ID in the config to scope queries to a specific Ragie partition.

コメント

「その他」の他のコンテンツ