What is Openbook?
The mental model behind Openbook — organizations, spaces, and rooms — and where the AI fits in.
Openbook is a team workspace built from three layers. Once you understand how they fit together, everything else in the product falls into place.
The three layers
Organization → Space → Room. That's the whole hierarchy.
An organization is the top-level container — usually your company. Everyone you work with joins the organization first, and one account can belong to several organizations.
Organizations are not Openbook's, they are your Bookbag account's. The same organization is the one you have in every Bookbag product, and who is in it, what role each person holds, who owns it and who has been invited are all decided at the account service. Openbook reads that list; it does not keep its own. Billing for Openbook, on the other hand, is Openbook's — each organization has its own Openbook plan. See Roles and permissions.
A space is a home for a team or a project inside the organization. Your engineering team might have a space, your marketing team another, and a big cross-functional launch a third. A space can be visible to everyone in the organization or restricted to only certain people.
A room is where the actual work happens. Each space holds as many rooms as you need, and you add them from the space sidebar at any time.
A room is an app, not a folder
This is the part that surprises people coming from folder-based tools. A room isn't a container for files — it's a fully-built, purpose-specific app that lives inside your space. When you add a kanban room, you get a working kanban board. When you add a docs room, you get a document editor. Nothing to install, nothing to configure elsewhere.
There are 21 room types: feed, project status, table board, kanban, AI agent, Q&A, discussions, check-in, docs, wiki, dashboard, whiteboard, gantt, retrospective, group, event, chat, proofing, feedback, time off, and town hall — covering planning, writing, discussion, and tracking without leaving the space.
You don't have to choose room types up front. Spaces start with whatever you pick (or whatever a template provisions), and you can add or remove rooms whenever the team's needs change. For guidance on which types fit which jobs, see Rooms → Choosing your rooms.
A worked example
Say your company is "Acme". You'd have:
- Organization: Acme — everyone at the company is a member here, and this is where billing lives.
- Space: Product Team — visible to the whole org.
- Rooms inside it: a kanban board for the sprint, a docs room for specs, a chat room for quick questions, and a retro room for end-of-sprint reviews.
A second space — say, "Website Redesign" — might be restricted to just the five people working on it, with its own boards and docs that the rest of the org never sees.
Where the AI fits
Openbook's AI shows up in two ways:
- The assistant is available throughout the workspace. Ask it questions, have it summarize activity, or let it act on your work — it operates in the context of whatever you're looking at.
- The build agent lives in its own room type and works against a real repository: you describe what you want, it edits files and opens a pull request. See AI → Build agent.
- The board MCP server points the other way: your own Claude or ChatGPT connects to a Kanban board and works it directly. See AI → Board MCP server.
Both are included during your organization's Pro trial; on the Free plan the assistant is limited to a monthly allowance of actions. Plan details live in Billing → Plans.
What to do next
- Create your workspace — sign in, create an organization, and set up your first space.
- Invite your team — Openbook is built for groups; it gets useful fast once teammates are in.
- Skim Roles and permissions so you know who can see and edit what before you open things up.