QuestionSetExecution

A decision service that answers every question of a request in one provider operation.

The provider receives the complete question set rather than one question at a time. A prompted LLM service can implement this with one prompt containing all the questions. This hook does not select the model's structured-output mode.

Execution sends the whole request to this method when the service implements it. The service's capabilities list every question kind it answers this way. Execution reaches this method only after the request's question kinds have passed preflight. A provider may still reject provider-specific limits before sending the request.

Inheritors