I understand that Assist does not have full session detail information at this point but what i would like to do is have some level of understanding of past message or messages and adjust my prompt. Is this possible? I have seen some demos where people can for instance move my desk up and then after say ok, now move "it" down and it understands but I have not been able to replicate.
In case it matters, I am using an LLM based coversation process and not just basic assist.