Chain of Thought MCP Server
Anthropic's recent article ["The "think" tool: Enabling Claude to stop and think in complex tool use situations"](https://www.anthropic.com/engineering/claude-think-tool) shows that using an external `think` tool notably increases performance on [SWE Bench](https://www.swebench.com/).
This MCP Server uses Groq's API to call LLMs which expose raw chain-of-thought tokens from Qwen's [qwq model](https://qwenlm.github.io/blog/qwq-32b/).
Installation
- Clone this repository to your local machine.
- Run ```uv sync``` to install depencies
- Get a Groq API key from [here](https://console.groq.com/keys).
- Update your mcp configuration with:
The path should be the local path to this repository. You can get this easily by running `pwd` in the terminal from the root of the repository.
Instructing The AI To Use This MCP Server
I personally prefer the agent call this tool on every request to increase performance. I add this to my rules for the agent: