I just use my memory and verify later. Unlike a LLM I have persistent long term durable storage of knowledge. Typically I can pretty easily pick out a hallucination though because there’s often a very clear inconsistency or logical leap that is non sense.