The popular local LLM inference engine llama.cpp now fully supports the Model Context Protocol (MCP), enabling standardized tool integration for local models. The update was discussed on Reddit's r/LocalLLaMA.