Add the result to the conversation context, and prompt to generate output without interrupting ongoing generation.