Chain Of Thought

Pain Points and Solutions

Chain Of Thought is designed to address the three core pain points of AI programming assistants in task management.

Memory Loss

AI assistants lack the ability to remember tasks across conversations, leading to difficulty in tracking long-term task progress, repeating explanations of the same needs, wasting time and resources.

Task Memory Function

Automatically saves execution history, provides long-term memory, allowing AI assistants to remember previous task progress, seamlessly continuing unfinished tasks.

Structure Chaos

Complex tasks lack systematic management, leading to low efficiency, lack of dependency management, chaotic sub-task execution, and difficulty in tracking overall progress.

Structured Task Decomposition

Automatically decomposes complex tasks into manageable sub-tasks, establishes clear dependency relationships, provides an orderly execution path, ensuring efficient completion.

Repetitive Work

Unable to effectively utilize past experiences and solutions, starting from scratch each time, lacking knowledge accumulation and reference system.

Knowledge Accumulation and Experience Reference

Automatically records successful solutions, establishes a task knowledge base, supports quick reference for similar tasks, and achieves knowledge accumulation and reuse.

Explore Core Features

Core Features

Chain Of Thought provides six core features to help you manage, execute, and track complex tasks efficiently.

Intelligent Task Planning and Analysis

Through in-depth analysis of requirements and constraints, generate structured task plans. Automatically assesses scope, risks, and priorities, providing rational and comprehensive implementation strategies.

Automatic Task Decomposition and Dependency Management

Intelligently break down complex tasks into manageable smaller tasks, identify dependencies between tasks, establish optimized execution paths, and avoid resource conflicts and execution bottlenecks.

Execution Status Tracking

Monitor the execution status of each task in real time, provide visual progress display, automatically update dependency status, and provide detailed execution reports upon task completion.

Task Integrity Verification

Comprehensively check task completion, ensure all requirements and standards are met, provide verification reports and quality assessments, and ensure deliverables meet expected requirements.

Task Complexity Assessment

Assess task complexity based on multi-dimensional criteria, provide resource requirement estimates, identify high-risk components, and help allocate resources and time reasonably.

Task Memory Function

Provide cross-session task memory, automatically save execution history and context, allow resuming and continuing tasks at any time without repeating requirements.

Learn the Workflow

Workflow

Chain Of Thought provides a complete workflow, with each step from task planning to completion carefully designed.

Task Planning

Initialize and plan the task process in detail

In-depth Analysis

Analyze requirements in depth and assess technical feasibility

Solution Reflection

Critically review analysis results and optimize the solution

Task Decomposition

Break down complex tasks into manageable sub-tasks

Task Execution

Execute specific tasks according to the scheduled plan

Result Verification

Comprehensively verify task completion and quality

Task Completion

Mark task as completed and generate report

Task Planning

Initialize and plan the task process in detail, establish clear goals and success criteria, and choose to continue planning based on existing tasks.

Learn More →

In-depth Analysis

Analyze requirements in depth and assess technical feasibility and potential risks, providing preliminary solutions.

Learn More →

Solution Reflection

Critically review analysis results, assess the completeness of the solution, and identify opportunities for optimization to ensure the solution meets best practices.

Learn More →

Task Decomposition

Break down complex tasks into independent and trackable sub-tasks, establish clear dependencies and priorities, and support multiple update modes.

Learn More →

Task Execution

Execute specific tasks according to the scheduled plan, ensuring that each step meets the quality standards and handling exceptions in the execution process.

Learn More →

Result Verification

Verify the completion of the task and ensure all requirements and technical standards are met, providing a quality assessment report without missing details.

Learn More →

Task Completion

Mark the task as completed and generate a detailed completion report, update the dependency status of related tasks, and ensure the continuity of the workflow.

Learn More →

Task Planning and Decomposition Process

This example demonstrates how to use MCP Chain of Thought to plan and decompose complex tasks. The entire process includes four main steps:

Task Planning - Initialize and plan the task in detail, clarify goals and success criteria
In-depth Analysis - Gain a deep understanding of the task, analyze technical feasibility and potential challenges
Solution Reflection - Critically review analysis results and optimize the proposal
Task Decomposition - Break down complex tasks into manageable sub-tasks

With this approach, you can transform complex large tasks into structured, actionable work units while maintaining an overall perspective.

Task Execution and Completion Process

This example demonstrates how to execute and complete planned tasks. The entire process includes four main steps:

Task List - Query the list of pending tasks to understand the current status
Task Execution - Execute the selected task according to the scheduled plan
Result Verification - Verify task completion and ensure quality standards are met
Task Completion - Formally mark the task as completed and generate a report

With this approach, you can systematically execute tasks and ensure each step meets the expected quality standards, ultimately completing the entire workflow.

💡 Tip

The above workflow is not fixed; the Agent will iteratively repeat different steps based on the analysis until the expected result is achieved.

Prompt Customization Feature

Customize system prompts via environment variables, allowing you to tailor AI assistant behavior without modifying code

Feature Overview

Prompt customization allows users to adjust AI assistant behavior via environment variables, offering two customization methods: complete override of the original prompt or appending content to the existing prompt.

Main Benefits

Personalization: Adjust system behavior according to specific project or domain needs
Efficiency Improvement: Optimize for repetitive task types, reducing redundant explanations
Brand Consistency: Ensure output content aligns with organizational style guides and standards
Professional Adaptability: Adjust terminology and standards for specific technical fields or industries
Team Collaboration: Standardize prompts used by team members to ensure consistent workflows

Usage

Environment Variable Configuration

Set environment variables to customize prompts for each feature, using specific naming conventions:

# Override Mode: MCP_PROMPT_[FUNCTION_NAME]
# Append Mode: MCP_PROMPT_[FUNCTION_NAME]_APPEND

# Example:
MCP_PROMPT_PLAN_TASK="Custom task planning prompt"
MCP_PROMPT_EXECUTE_TASK_APPEND="Additional task execution guidance"

See the detailed documentation for more configuration methods and parameter usage instructions.

View Full Documentation

Installation and Configuration

Chain Of Thought provides multiple installation methods, making it easy to get started quickly or perform advanced configuration as needed.

Manual Installation Setup

1

Clone the Repository

git clone https://github.com/liorfranko/mcp-chain-of-thought.git
cd mcp-chain-of-thought

2

Install Dependencies

npm install

3

Build the Project

npm run build

Cursor IDE Configuration

If you use Cursor IDE, you can integrate Chain Of Thought into your development environment.

Your project directory/.cursor/mcp.json

{
  "mcpServers": {
    "mcp-chain-of-thought":{
      "command": "node",
      "args": [
        "/path/to/mcp-chain-of-thought/dist/index.js"
      ],
      "env": {
        "DATA_DIR": "/path/to/project/data"
      }
    }
  }
}

Important Note

You must use absolute paths: Please ensure that the DATA_DIR configuration uses absolute paths instead of relative paths, otherwise data may not load correctly.

Prompt Configuration Instructions

Chain Of Thought supports two modes:

TaskPlanner: Suitable for initial task planning and complex task decomposition, the AI assistant acts as a task planner.


You are a professional task planning expert. You must interact with the user, analyze their needs, and collect project-related information. Finally, use "plan_task" to create tasks. After the task is created, you must summarize and inform the user to use the "TaskExecutor" mode for task execution.
You must focus on task planning and are strictly prohibited from using "execute_task" to execute tasks.
Warning: You are a task planning expert. You cannot directly modify code. You can only plan tasks, and you cannot directly modify code. You can only plan tasks.

TaskExecutor: Suitable for executing predefined tasks, the AI assistant acts as an execution expert.


You are a professional task execution expert. When the user specifies a task to execute, use "execute_task" to perform the task.
If no task is specified, use "list_tasks" to find unexecuted tasks and execute them.
After execution, you must summarize and inform the user of the conclusion.
You can only execute one task at a time. Unless the user explicitly requests, do not proceed to the next task.
If the user requests "continuous mode," execute all tasks in order.

You can use the Custom modes configuration in Cursor settings to define modes for different work scenarios.

Quick Start

After installation, check out our quick start guide to learn how to use MCP Chain of Thought.

View Code →

FAQ

Have questions? Check our FAQ or submit an issue on GitHub.

View FAQ →

Pain Points and Solutions

Memory Loss

Task Memory Function

Structure Chaos

Structured Task Decomposition

Repetitive Work

Knowledge Accumulation and Experience Reference

Core Features

Intelligent Task Planning and Analysis

Automatic Task Decomposition and Dependency Management

Execution Status Tracking

Task Integrity Verification

Task Complexity Assessment

Task Memory Function

Workflow

Task Planning

In-depth Analysis

Solution Reflection

Task Decomposition

Task Execution

Result Verification

Task Completion

Task Planning

In-depth Analysis

Solution Reflection

Task Decomposition

Task Execution

Result Verification

Task Completion

Task Planning and Decomposition Process

Task Execution and Completion Process

💡 Tip

Prompt Customization Feature

Feature Overview

Main Benefits

Usage

Environment Variable Configuration

Installation and Configuration

Manual Installation Setup

Clone the Repository

Install Dependencies

Build the Project

Cursor IDE Configuration

Important Note

Prompt Configuration Instructions

Quick Start

FAQ

Experience Intelligent Task Management Now