Contextual Chat Feature

[jnsahaj]

I've been thinking of a chat feature for lumen which would remember the context of the previous message, instead of just generating the summary.

I would like to gather feedback on how the feature should be expected to work.

For example, from what I can think of now, it can be as simple as

lumen explain HEAD --chat
>> <AI Summary>
>> What exactly changed in the header component?
>> <AI Response>
>> Okay, can you suggest a better way to do this?
>> <AI Response>