Builds an LLMOneShotCallback that calls a server-side LLM to infer a PK. Returns null if AIEngine isn't loadable or no language model is active — pipeline silently skips the LLM tier in that case (verdict cascades to none).
LLMOneShotCallback
null
none
Builds an
LLMOneShotCallbackthat calls a server-side LLM to infer a PK. Returnsnullif AIEngine isn't loadable or no language model is active — pipeline silently skips the LLM tier in that case (verdict cascades tonone).