MCP Reasoner

A systematic reasoning MCP server implementation for Claude Desktop with beam search and thought evaluation.

typescript

277

GitHub Stars

typescript

Language

4 months ago

First Indexed

3 weeks 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": {
    "jacck-mcp-reasoner": {
      "command": "node",
      "args": [
        "path/to/mcp-reasoner/dist/index.js"
      ]
    }
  }
}

You can add the MCP Reasoner to Claude Desktop to enable advanced problem-solving capabilities. It provides two reasoning strategies—Beam Search and Monte Carlo Tree Search (MCTS)—and helps you analyze and map Claude’s thinking through complex problems for better control and insight.

How to use

Integrate the MCP Reasoner with Claude Desktop to let Claude explore problems using either Beam Search for straightforward reasoning or MCTS for complex scenarios. You can switch between the two strategies depending on the task, and you will gain visibility into how different reasoning paths are evaluated and how the final conclusions were reached.

How to install

Prerequisites you need on your system:

Node.js installed (the MCP server runs as a Node process).
Access to a terminal or command prompt. Follow these concrete steps to install and prepare the MCP Reasoner:

git clone https://github.com/frgmt0/mcp-reasoner.git

OR clone the original:

git clone https://github.com/Jacck/mcp-reasoner.git

cd mcp-reasoner
npm install
npm run build

Additional notes

Configure Claude Desktop to use the MCP Reasoner by adding a new MCP server entry in your Claude config. The following configuration shows how to point Claude to the local built script.

{
  "mcpServers": {
    "mcp-reasoner": {
      "command": "node",
      "args": ["path/to/mcp-reasoner/dist/index.js"]
    }
  }
}

Available tools

beam_search

A reasoning method that tracks multiple promising paths and expands the most favorable ones step by step to reach a solution with efficient exploration.

mcts

A Monte Carlo Tree Search variant used for exploring complex decision spaces by simulating many possible futures and selecting paths with high expected value.