Manages a local fleet of llama.cpp GGUF models, enabling AI agents to discover, start, stop, test, and query evaluation results for llama-server processes.
