Expand description
llama.cpp provider: a local, OpenAI-compatible server managed by vtcode.
Module layout (per crate-local AGENTS.md: providers/<name>/mod.rs):
mod.rs(this file):LlamaCppProvider, provider trait impls, request model selection, public API surface.probe.rs: HTTP model discovery (fetch_llamacpp_models),ServerProbe,probe_server.startup.rs: model-path policy, binary/args resolution, process spawning, readiness polling.managed.rs: shared state registry, phase tracking, readiness orchestration (ensure_server_ready).