Why Can’t Rovo Return the Users of a Confluence Space? [Champions Slack Insider]

[#1]

This thread sparked by @Ciara Twomey Nielsen gets at a very practical (and slightly frustrating) question:

Why can’t Rovo return the users of a space—even when I can see them in the UI?

What’s actually happening

Short Answer: Space membership is not a first-class, queryable object in the Teamwork Graph today.

Even though Confluence shows you all users in a space, that view is not coming from a single dataset that Rovo can query. It is assembled at runtime based on permissions, groups, and access rules.

Rovo can only reason over:

• entities explicitly modeled in the Teamwork Graph
• relationships that have been defined and exposed

Space membership, as a direct “listable” object, is not fully exposed in that way yet.

Why the UI shows it but Rovo can’t

This is the key disconnect:

• The Confluence UI can display space users because it evaluates permissions dynamically
• Rovo depends on indexed, structured graph data

So even if something is visible on screen, that does not mean it exists as a queryable object in the graph.

What about macros or page content?

Another nuance raised in the thread:

Rovo does not “see” the page the way a user does. It works from stored content and structured data. Some macros or dynamic views do not translate cleanly into something Rovo can interpret.

So even if a page visually lists users, that does not guarantee Rovo can extract or reason over that list.

Why does Rovo still say “Project” instead of “Space”?

Many parts of Rovo are still influenced by Jira-first modeling. Rovo does understand Confluence spaces conceptually, but:

• terminology
• query patterns
• underlying models

often default to project-based language.

This is a known gap, but there is no published timeline for full alignment.

Can you see what’s in the Teamwork Graph?

Not in a complete, user-friendly way. Today, visibility is fragmented across:

• connector documentation
• supported features
• developer documentation

There is no full graph explorer or single interface where you can browse everything available.

One practical option is:

• Rovo debug view, which shows what entities were retrieved for a query

But even that is partial.

Can customers define their own objects?

Short Answer: Not fully.

There is developer-level documentation describing:

• object types
• relationships
• how data flows in

But:

• customers cannot freely define new graph object types
• the graph is not fully open or extensible yet

Expansion is happening gradually, with governance and privacy considerations slowing things down.

Champion takeaway

Rovo does not operate on everything you can see in the UI. It operates on what has been explicitly modeled in the Teamwork Graph.

If something is:

• dynamically generated
• permission-derived
• or not exposed as a structured object

Rovo will struggle to return it directly.

If you’ve ever thought, “If I can see it, Rovo should be able to,” this is one of the clearest examples of why that assumption does not always hold.

The gap is not visibility. It is data modeling.