Sets up, manages, and optimizes locally-hosted LLMs (Ollama, LMStudio, llama.cpp) with detection, configuration, benchmarking, and model selection capabilities.