Skip to main content
A shared brain is a memory your team builds together: the company knowledge base, a client project, a product area. A shared brain is simply a brain with more than one member; the member list decides who can see it and what each person can do.

Nothing arrives by accident

Nothing flows into a shared brain on its own. Every write names its destination brain explicitly, whether it comes from a person working in the inbox, an agent acting on someone’s instruction, or a connected tool pointed at that specific brain. An agent can only write to brains its person can write to, checked live on every call, and when someone captures without naming a destination, it goes to their own brain, not a shared one. For something to appear in a shared brain, someone with Write access put it there: directly, through their agent, or by promoting it from their own brain.

Members and roles

People are added to a brain with one of four roles: Your agents inherit these roles: an agent connected by a Reader can only read the brain, whoever asks it to do otherwise.

Creating one

Creating brains is a workspace admin action. From Brains, an admin creates a new brain, gives it a name and a description (write the description for your agents: it is how they decide what belongs there), picks a template that matches what it is for, and adds members. A brain for a new teammate can also be created automatically from their invite. It starts empty; its inbox works exactly like any other.

Try it

  1. Have a workspace admin create a shared brain called “Team Handbook” using the Company template.
  2. Add a teammate with the Write role.
  3. Drop a document into its inbox and watch it become the brain’s first page.