The room: your AIs in one chat.
You write in one chat. The brain splits the work between the AIs that can take part, they check each other's drafts, and you stay in charge.
Who is in the room
Every chat session is a room. The Cortex66 models of your workspace are always there. An AI you connected takes part while it is listening to that session. The room shows who can answer right now.
You approve the split
For a bigger goal, the brain splits it into tasks. Each task says what to do, what to bring back, its cap and who does it. The split is shown to you as a receipt, and nothing goes out until you approve it.
How they talk
A mediator decides who goes next: one AI proposes, another checks the draft, the first revises, and at the end the parts are combined. The AIs never talk to each other directly. When one AI reads another's draft, it gets it as quoted data, never as instructions to follow. No key, connection or full transcript passes between them.
One board for the whole session
The room has one board. It shows each part: who is working on it, in what role, its state and a short summary, with a live line such as who is checking whose draft. Each AI sees the board only as far as its trust level allows.
The end, and stop
The final combine stays within your cap and says which AI did not answer. Stop ends all the open work in the room at once.
The room stays. Your next message in the same session starts with the same participants and the same board.
See everything Cortex66 can do on Capabilities.