MCP Chain of Thought

> 🚀 An intelligent task management system based on Model Context Protocol (MCP), providing an efficient programming workflow framework for AI Agents.

📑 Table of Contents

• ✨ Features
• 🧭 Usage Guide
• 🔧 Installation
• 🔌 Using with MCP-Compatible Clients
• 🛠️ Tools Overview
• 🤖 Recommended Models
• 📄 License
• 📚 Documentation

✨ Features

• 🧠 Task Planning & Analysis: Deep understanding of complex task requirements
• 🧩 Intelligent Task Decomposition: Break down large tasks into manageable smaller tasks
• 🔄 Dependency Management & Status Tracking: Handle dependencies and monitor progress
• ✅ Task Verification: Ensure results meet requirements
• 💾 Task Memory: Store task history for reference and learning
• ⛓️ Thought Chain Process: Step-by-step reasoning for complex problems
• 📋 Project Rules: Define standards to maintain consistency
• 🌐 Web GUI: Optional web interface (enable with `ENABLE_GUI=true`)
• 📝 Detailed Mode: View conversation history (enable with `ENABLE_DETAILED_MODE=true`)

🧭 Usage Guide

🚀 Quick Start

1. 🔽 Installation: Install MCP Chain of Thought via Smithery or manually

2. 🏁 Initial Setup: Tell the Agent "init project rules" to establish project-specific guidelines

3. 📝 Plan Tasks: Use "plan task [description]" to create a development plan

4. 👀 Review & Feedback: Provide feedback during the planning process

5. ▶️ Execute Tasks: Use "execute task [name/ID]" to implement a specific task

6. 🔄 Continuous Mode: Say "continuous mode" to process all tasks sequentially

🔍 Memory & Thinking Features

• 💾 Task Memory: Automatically saves execution history for reference
• 🔄 Thought Chain: Enables systematic reasoning through `process_thought` tool
• 📋 Project Rules: Maintains consistency across your codebase

🔧 Installation

🔽 Via Smithery

npx -y @smithery/cli install @liorfranko/mcp-chain-of-thought --client claude

🔽 Manual Installation

npm install
npm run build

🔌 Using with MCP-Compatible Clients

⚙️ Configuration in Cursor IDE

Add to your Cursor configuration file (`~/.cursor/mcp.json` or project-specific `.cursor/mcp.json`):

{
  "mcpServers": {
    "chain-of-thought": {
      "command": "npx",
      "args": ["-y", "mcp-chain-of-thought"],
      "env": {
        "DATA_DIR": "/path/to/project/data", // Must use absolute path
        "ENABLE_THOUGHT_CHAIN": "true",
        "TEMPLATES_USE": "en",
        "ENABLE_GUI": "true",
        "ENABLE_DETAILED_MODE": "true"
      }
    }
  }
}

> ⚠️ Important: `DATA_DIR` must use an absolute path.

🔧 Environment Variables

• 📁 DATA_DIR: Directory for storing task data (absolute path required)
• 🧠 ENABLE_THOUGHT_CHAIN: Controls detailed thinking process (default: true)
• 🌐 TEMPLATES_USE: Template language (default: en)
• 🖥️ ENABLE_GUI: Enables web interface (default: false)
• 📝 ENABLE_DETAILED_MODE: Shows conversation history (default: false)

🛠️ Tools Overview

🤖 Recommended Models

• 👑 Claude 3.7: Offers strong understanding and generation capabilities
• 💎 Gemini 2.5: Google's latest model, performs excellently

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📚 Documentation

• 🏗️ System Architecture
• 🔧 Prompt Customization Guide
• 📝 Changelog

⭐ Star History