Rovo can summarize Confluence content and changes, while authors can record version comments and maintain a deliberate summary. This comparison tests both baselines before a addly macro concept advances.
Choose Rovo when its page or change summary answers the reading question. Choose an author-written block when accountability and exact wording matter more than generation.
Page Summary is a concept macro with no connected AI service, distributed version, Marketplace listing, installation test, verified data flow, price, or support commitment. The shown extraction behavior is fictional.
Test the same representative page, the same reader question, and the same acceptance criteria. Generated brevity is not the same as an accountable decision record.
Official documentation describes page, live-doc, and blog summaries, plus change summaries since a visitor's last visit. Atlassian also documents limitations and access conditions that must be checked in the target organization.
The page owner can maintain a short decision-and-actions section and use Confluence version comments to describe a published update in accountable language.
The hypothesis is a macro that presents cited decisions, actions, owners, and open questions in a stable structure with human validation. No extraction pipeline currently exists.
Blocked: concept only, no model, data flow, or verified output
The right answer depends on the job: understand the page, understand recent change, or publish a controlled decision record.
| Criterion | Native baseline | Manual baseline | Concept hypothesis | Decision rule |
|---|---|---|---|---|
| Page orientation | Rovo can generate a summary of a page, live doc, or blog when sufficient acceptable content is available. | The author writes the exact executive summary readers should rely on. | Would offer a consistent macro structure with citations and explicit validation. | Keep Rovo when orientation—not authoritative publication—is the only job. |
| Recent changes | Rovo can summarize changes since the visitor's last visit; Atlassian states that macros, images, and links are currently excluded. | Version comments describe the intent of each published update for watchers and history. | Would need a defined comparison window and transparent handling of unsupported content. | Keep native when the documented omissions do not affect the decision. |
| Accountability | Generated output can be rated, but quality, accuracy, and reliability require human judgment. | A named owner controls wording and signs the review date. | Would require human approval, source anchors, edit history, and a non-generated fallback. | Prefer manual when the summary itself is a controlled record. |
| Data and operating risk | Uses the organization's configured Confluence and Rovo environment. | Adds editorial effort without another processing service. | Would add permissions and potentially a processor, retention, model, compatibility, and support obligations. | Reject the app path until its real architecture and data flow are published. |
A generated-looking macro is not evidence of reliable summarization.
A summary may not generate for insufficient content or acceptable-use concerns; access and availability also depend on the target organization.
Atlassian currently states that macros, images, and links are not included in the change summary.
The demonstration does not process real page content and proves no accuracy, completeness, latency, or data-handling property.
Native claims and limitations are taken from current Atlassian Support documentation. Confirm organization settings and the latest terms before evaluation.
Reviewed
Official Atlassian source
Summary flow, feedback, insufficient-content behavior, access conditions, and Trust Center reference.
Open sourceOfficial Atlassian source
Since-last-visit behavior, page-summary fallback, copy and feedback actions, and excluded macros, images, and links.
Open sourceOfficial Atlassian source
Version comments, publishing, update history, change viewing, version comparison, and restoration.
Open sourceUse representative content and keep generated output clearly separate from approved truth.
Explore the intended macro with fictional decisions, actions, and owners.
Open demoInspect the intended review flow, permissions, checks, and unresolved limits.
Open documentationReview the absent processor, compatibility, verification, and support evidence.
Open trust profileEnter your page volume, review effort, costs, and realistic adoption.
Open ROI modelRecord the corpus, expected answer, exclusions, sources, reviewer, and stop conditions.
Open evaluation roomOnly continue when a stable, cited, human-approved structure solves a measured gap—and a real implementation can be tested for accuracy, omissions, permissions, and data handling.