llm-helper engines

List the inference engines a profile or recipe can run on

Engines are the inference servers llm-helper can launch, such as vLLM or llama.cpp. Each is a definition: how it is installed, its command line, how it takes the model and the API key, and how its health is read. They ship in the binary, and files in /engines.d can add more without a rebuild. A profile picks one with PROVIDER, and a recipe with its provider field.

Usage

llm-helper engines