> ## Documentation Index > Fetch the complete documentation index at: https://docs.clarityq.ai/llms.txt > Use this file to discover all available pages before exploring further. # Unified Workspaces > Ask one question across several workspaces and get a single answer back. A **unified workspace** answers across several of your other workspaces at once. Instead of asking the same question in three places and reconciling the results yourself, you ask once and get one answer covering all of them. The workspaces it draws on are called its **member workspaces**. A unified workspace has no data warehouse of its own — it reads each member's Context Layer and queries each member's warehouse. Use it for portfolio questions: comparing retention across your games, rolling up revenue across business units, or spotting which product is pulling an org-wide number around. ## Choosing What a Question Spans A picker in the chat input shows the member workspaces a question will cover. Add or remove them before you ask. Two things to know: * **At least one member must stay selected.** There's also a limit on how many a single conversation can span. * **The selection locks once the conversation starts.** To ask across a different set, start a new chat. This keeps every answer in a conversation comparable with the ones before it. The members a conversation ran against are shown in its header, and travel with anything you save from it — so a dashboard or a shared chat always says which workspaces its numbers came from. ## Reading the Answer Every figure is attributed to the workspace it came from. The agent won't quietly add numbers together across workspaces: definitions differ between them, so it checks they're comparable before combining them and tells you when they aren't. Each member is queried on its own terms, using that workspace's definitions, its own warehouse, and its own skills and memory. ## What's Different Inside One A unified workspace sits on top of other workspaces rather than holding data itself, so parts of the product don't apply. **Its Context Layer holds two things**: [Skills](/context-layer/skills) and [Memory](/context-layer/memory). These are for knowledge that spans the portfolio — a methodology for comparing products, or a fact true across all of them. In the Builder you can create those two and nothing else. **There's no Table Catalog, Event Catalog, Semantic Catalog, or Files**, because those belong to the member workspaces. You can still see a member's components when the agent uses them, but read-only — to change one, go to the workspace that owns it. Skills and memory saved in a unified workspace are used only by its own chats. A member workspace doesn't inherit them, and a fact saved here won't apply when someone asks that workspace directly. ## Diagnosis and Context Coverage Both work here. The difference is that a gap belongs to whichever workspace owns the thing that's missing, and that's usually one of the members rather than the unified workspace itself. So when you act on a gap: * **Fixing** opens the Builder of the workspace that owns it. * **Reporting** files into that workspace's Inbox, where the people who maintain it will see it. * **What you can do** depends on your rights in that member workspace, which is why two rows in the same report can offer different options. Less often, the gap is in the unified workspace's own skill or memory — a cross-workspace methodology that's wrong or missing. Those behave normally: you fix them here, and reports go to this workspace's Inbox. See [Context Coverage](/ask-anything/context-coverage) and [Diagnosing Answers](/ask-anything/diagnosing-answers) for how both work generally. ## Dashboards A dashboard you build here works like any other, except that it carries the member workspaces it was built against. That scope is part of the dashboard: the numbers mean nothing without knowing which workspaces they cover, so it's shown alongside the dashboard and kept with it. That scope also decides who can open it. Someone needs access to **every** member workspace in the dashboard's scope, not just one of them. If they're missing any, the dashboard won't open and they're pointed at its owner to ask. This catches people out when sharing broadly: a dashboard spanning five workspaces is only openable by people who can reach all five, even though the dashboard itself may look no different from a single-workspace one. ## Who Can Use One There's no separate role, and nobody gains access to data they didn't already have. An organization has a single unified workspace, but it isn't the same view for everyone. **Each person reaches only the member workspaces they're assigned to.** The picker shows your members, so two colleagues in the same unified workspace can see different options and get answers covering different ground. An answer never spans a workspace you couldn't have asked directly. The same rule governs everything saved from it — dashboards, scheduled tasks, and shared chats all stay within the scope they were created with. ## Setting One Up A unified workspace isn't something you create yourself. Ask your ClarityQ contact, and tell them which workspaces it should span. They configure which workspaces are available to it, which of those are selected by default when someone starts a chat, and how many a single conversation may cover. After that it appears alongside your other workspaces, and people use it by switching to it. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.