Enables two local LLMs (Qwen and Llama) to communicate via MCP, allowing one model to consult the other for second opinions or additional reasoning.
