Roles and permissions

There are two layers of role in Cortex: your role in the workspace (what you can do to the workspace itself) and your role on each brain (what you can do inside it). Being in a workspace does not open its brains to you; you are added to each brain with a role that says what you can do there.

​

Workspace roles

Creating brains is an admin action. Everyone gets brains by being added to them: by an admin, by a brain’s own Admin or Owner, or automatically through their invite.

​

Brain roles

​

The questions people actually ask

Can a Reader’s agent write to the brain? No. Your agents act as you, with your roles, checked live on every call. A Reader’s agent can only read. Can a Writer delete pages? A Writer can create and edit pages, and edits keep full version history so nothing is truly lost. Managing the brain itself, including its members and settings, is an Admin and Owner responsibility. Who can change how much gets filed automatically? The brain’s Owner sets its filing mode. Can teammates see my work brain? Only if you add them. A brain is visible to its members and no one else, and your work brain starts with a member list of one: you. Workspace admins administer brains as part of running the workspace (for example when someone leaves), but they are not members of your brain and do not see it in daily use. What can a guest do? Exactly what their brain roles say, in the brains they were granted, and nothing more. Guests do not see the workspace’s team directory or its other brains.

​

Your agents inherit your roles

A connected agent acts as you: it sees the brains you can see and holds the role you hold on each, resolved live. If your role on a brain changes, every agent you have connected changes with it, immediately. More on connections.