Macro frontend for defining OpenAI-compatible upstream providers.
The generated provider implements LLMProxy.Providers.Behaviour and delegates
token selection, request execution, streaming, and usage extraction to the
shared OpenAI-compatible client.