drio
Open app

RAT MCP Server (Retrieval Augmented Thinking)

Source

🧠 MCP server implementing RAT (Retrieval Augmented Thinking) - combines DeepSeek's reasoning with GPT-4/Claude/Mistral responses, maintaining conversation context between interactions.

Catalog onlyCatalog onlySTDIO

Overview

RAT MCP Server (Retrieval Augmented Thinking) is a server that implements a two-stage reasoning process, combining DeepSeek's reasoning capabilities with various response models like GPT-4 and Claude, while maintaining conversation context.

To use the RAT MCP Server, clone the repository, install dependencies, configure your API keys in a .env file, and build the server. You can then integrate it with Cline for generating responses.

  • Two-stage processing using DeepSeek for reasoning and multiple models for response generation. - Maintains conversation context and history. - Supports various models including Claude and OpenRouter models.
  1. Enhancing AI responses through structured reasoning.
  2. Providing context-aware answers in conversational AI applications.
  3. Integrating with development tools for AI-assisted coding.

Add to your AI client

Use these steps to connect RAT MCP Server (Retrieval Augmented Thinking) in Cursor, Claude, VS Code, and other MCP-compatible apps. The same JSON appears in the Use with menu above for one-click copy.

Cursor

Add this to your .cursor/mcp.json file in your project root, then restart Cursor.

.cursor/mcp.json

{
  "mcpServers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

Claude Desktop

Add this server entry to the mcpServers object in your Claude Desktop config, then restart the app.

~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows)

{
  "mcpServers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

Claude Code

Add this to your project's .mcp.json file. Claude Code will detect it automatically.

.mcp.json (project root)

{
  "mcpServers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

VS Code (Copilot)

Add this to your .vscode/mcp.json file. Requires the GitHub Copilot extension with MCP support enabled.

.vscode/mcp.json

{
  "servers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

Windsurf

Add this to your Windsurf MCP config file, then restart Windsurf.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

Cline

Open Cline settings, navigate to MCP Servers, and add this server configuration.

Cline MCP Settings (via UI)

{
  "mcpServers": {
    "rat-retrieval-augmented-thinking-mcp-newideas99": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-rat-retrieval-augmented-thinking-mcp-newideas99"
      ]
    }
  }
}

FAQ

What models does RAT MCP Server support?

It supports DeepSeek, Claude, and any OpenRouter models like GPT-4.

Is there a license for RAT MCP Server?

Yes, it is released under the MIT License.

How do I maintain conversation context?

The server automatically maintains conversation history and includes it in the reasoning process.

Research and data