Think Tool MCP Server (Simplified)
LocalAbout Think Tool MCP Server (Simplified)
A simplified MCP server for AI agents that need a dedicated place to record structured reasoning during complex tasks. Its single `think` tool logs thoughts without retrieving information or changing external state.
Key features
- 01Logs a supplied thought
- 02Does not retrieve information
- 03Does not change external state
Use cases
- 01Record reasoning during complex tasks
- 02Brainstorm without invoking external actions
MCP support
Exposes a single `think` tool to MCP clients. The tool accepts a required `thought` string and logs it without external effects.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEThink Tool MCP Server (Simplified)
A Model Context Protocol (MCP) server implementing the think tool as described by Anthropic. This server provides a dedicated space for an AI agent to perform structured thinking during complex tasks, improving reasoning and decision-making.
Overview
This simplified MCP server provides only the think tool. It allows an AI agent to log its reasoning process without affecting external state or retrieving new information. This implementation is based on the concept outlined in Anthropic's engineering blog post on enhancing agentic tool use. This specific think tool uses the adapted SWE-Bench version.
It is designed for use with Claude Desktop or any MCP-compatible client.
Available Tool
The server provides the following tool:
- think: Use the tool to think about something. It will not obtain new information or make any changes to the repository, but just log the thought. Use it when complex reasoning or brainstorming is needed.
- Parameters:
thought(required, string): A thought to think about.
- Input Schema:
{ "type": "object", "properties": { "thought": { "type": "string", "description": "Your thoughts." } }, "required": ["thought"] }
- Parameters:
Setup
Local Development
-
Clone the repository:
git clone https://github.com/monotykamary/think-tool-mcp.git cd think-tool-mcp -
Install dependencies:
npm install -
Build the project:
npm run build -
Run the server:
node dist/index.js
Environment Variables
No environment variables are required for the Think Tool MCP server.
Running with Docker
-
Build the Docker image:
docker build -t think-tool-mcp . -
Run the container:
docker run -i --rm think-tool-mcp
Usage with Claude Desktop
To use this MCP server with Claude Desktop, add the following configuration to your Claude Desktop settings:
{
"mcpServers": {
"think-tool": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"ghcr.io/monotykamary/think-tool-mcp"
]
}
}
}
With this configuration:
- Claude will automatically start the Docker container when needed.
- The MCP server will be available for logging and reviewing thoughts.
- The container will be removed when Claude Desktop is closed.
Usage Example
Logging a Thought
// Example MCP Tool Call
{
"tool_name": "think",
"arguments": {
"thought": "Okay, the user wants to refactor the database schema. First, I need to analyze the current schema (using get_schema tool), then identify potential improvements like normalization or indexing. After that, I'll propose a new schema and ask for confirmation before generating the migration script."
}
}
Troubleshooting
- Server Connection: Ensure the server is running (locally or via Docker) and configured correctly in your MCP client (e.g., Claude Desktop settings).
- Tool Name: Verify you are calling the tool named
think.
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.