Hey, thanks for coming back with more detail on the hub model. One coordinator as the single point of contact, with specialists working behind the scenes, is a clear and intentional pattern, not a messy chat artifact.
On your earlier question from #26: when you use a new group chat with the same Bot, the old 1:1 transcript isn’t pulled into the working set. Mohit confirmed this in a nearby thread: Grok Bot Ultra users: how do you make the weekly allowance last? Mine reached 99% in three days - #13 by mohitjain. So a new chat really starts with clean context, and right now it’s the closest way to get a thin working set without Duplicate or Delete. Also, durable state (profile, memory, routines, skills, connectors, files on a shared computer) lives separately from the transcript and is preserved.
That said, you’re right in principle. This is a workaround, not a product solution. First class controls like same Bot new session with auto archiving the old thread, a manual Compact that drops stale messages from the working set, and a visible working set meter aren’t shipped yet. Auto summarize at the context window boundary doesn’t replace that. All of this is on the team’s radar. I passed along your details, including the latency aspect, since response time grows with transcript size, not just cost. I can’t share a timeline for these controls yet, but if there’s an update I’ll post it here.