Multiplex a list-models operation onto the text-completion request/response
channel via an operation field (defaults to "completion" for backward
compatibility). Backends that support model listing return a normalised
ModelInfo with id, name, context limits, modalities, pricing, and local
runtime metadata. Unsupported backends return a not-implemented error.

Implemented for: OpenAI, Claude, Ollama, Mistral, vLLM, TGI, LlamaFile,
LM Studio, Azure OpenAI, Bedrock, VertexAI, Google AI Studio.

Adds tg-list-models CLI command with table and --json output.