GBrain gives a team, and every AI it uses, one shared memory. The documentation at gbrain.io/docs now comes with it.
Ask an AI connected to a workspace how to bring notes in from a computer, invite somebody, connect an account or set up a terminal, and it reads the page that answers the question and hands you the link. Before this it answered from what it already knew about GBrain, which was sometimes nothing and sometimes out of date.
The same is true in the workspace’s own chat, so a question about how GBrain works is answered the same way in ChatGPT, in Claude and in the chat.
The answer names the page it came from
Every answer comes back with the page it read, so the steps can be checked and shared. Ask how to move a memory into a workspace and the answer is the page on doing that, with its list of five steps.
Reading the docs costs nothing
The AI reading the documentation reaches no account of anyone’s and spends no credit. What it uses is the answer in the conversation, the same as any other question.
What you might notice
An AI that used to say it did not know how something in GBrain works now answers with the steps and a link. An AI that used to invent steps now has the real ones.
What you need to do
Nothing. Every AI connected to a workspace can read the documentation already, and there is no setting for it.
The documentation is at gbrain.io/docs, and What a connected AI can reach says what an AI gets with a connection.