Atlassian Release Notes

Jira Software release notes provide information on the features and improvements in each release. This page includes release notes for platform releases and feature releases (you'll find bug fix release notes after opening one of the versions below).

Upgrade matrix

Too many release notes? Take a look at our Upgrade matrix to get a quick roll-up of the most important changes in the latest versions.

Upcoming bug fix releases

Our bug fix releases follow a predefined, monthly schedule. This makes it easier for you to plan your next upgrade.

New bug fix versions are released for three Jira feature versions: Latest version, Jira 11.3 LTS, and 10.3 LTS. We may occasionally skip a bug fix release if it isn’t needed. All bug fix releases include security and regular bug fixes.

Bug fixes are released every month between the first Tuesday and the second Wednesday.

Jira Software Data Center 11 release notes

• Jira Software 11.3.x release notes LATEST LONG TERM SUPPORT
  • Latest bug fix release: 11.3.6
• Jira Software 11.2.x release notes
  • Latest bug fix release: 11.2.1
• Jira Software 11.1.x release notes
  • Latest bug fix release: 11.1.1
• Jira Software 11.0.x release notes
  • Latest bug fix release: 11.0.1

Jira Software Data Center 10 release notes

• Jira Software 10.7.x release notes
  • Latest bug fix release: 10.7.4
• Jira Software 10.6.x release notes
  • Latest bug fix release: 10.6.1
• Jira Software 10.5.x release notes
  • Latest bug fix release: 10.5.1
• Jira Software 10.4.x release notes
  • Latest bug fix release: 10.4.1
• Jira Software 10.3.x release notes LONG TERM SUPPORT
  • Latest bug fix release: 10.3.21
• Jira Software 10.2.x release notes
  • Latest bug fix release: 10.2.1
• Jira Software 10.1.x release notes
  • Latest bug fix release: 10.1.2
• Jira Software 10.0.x release notes
  • Latest bug fix release: 10.0.1

Long Term Support releases

An Atlassian Long Term Support release is a feature release that gets backported critical security updates and critical bug fixes during its entire two-year support window. If you can only upgrade once a year, consider upgrading to an LTS release.

Learn more

Developer releases

If you are looking for the release notes for the latest (Early Access Program) release, see Jira releases (Jira developer documentation) instead.

Earlier releases

Check the list of earlier Jira releases...

Last modified on May 18, 2026

Was this helpful?

Yes

No

Provide feedback about this article

In this section

• Jira Software 11.3.x release notes
• Jira Software 11.2.x release notes
• Jira Software 11.1.x release notes
• Jira Software 11.1.x upgrade notes
• Jira Software 11.0.x release notes
• Jira Software 11.0.x upgrade notes
• Jira Software 10.7.x release notes
• Jira Software 10.7.x upgrade notes
• Jira Software 10.6.x release notes
• Jira Software 10.6.x upgrade notes
• Jira Software 10.5.x release notes
• Jira Software 10.5.x upgrade notes
• Jira Software 10.4.x release notes
• Jira Software 10.4.x upgrade notes
• Jira Software 10.3.x release notes
• Jira Software 10.3.x upgrade notes
• Jira Software 10.2.x release notes
• Jira Software 10.2.x upgrade notes
• Jira Software 10.1.x release notes
• Jira Software 10.1.x upgrade notes
• Jira Software 10.0.x release notes
• Jira Software 10.0.x upgrade notes
• Jira Software 9.17.x release notes
• Jira Software 9.17.x upgrade notes
• Jira Software 9.16.x release notes
• Jira Software 9.16.x upgrade notes
• Jira Software 9.15.x release notes
• Jira Software 9.15.x upgrade notes
• Jira Software 9.14.x release notes
• Jira Software 9.14.x upgrade notes
• Jira Software 9.13.x release notes
• Jira Software 9.13.x upgrade notes
• Jira Software 9.12.x release notes
• Jira Software 9.12.x upgrade notes

Older

Related content

• Creating release notes
• JIRA Release notes documentation links take you to the wrong versions
• Creating release notes
• Creating release notes
• Jira Data Center automation release notes
• Jira build information
• Latest updates
• Jira Software documentation
• Release Notes page is Blank in Project Summary
• What are releases in your plan?

Powered by Confluence and Scroll Viewport.

Original source

Since we first launched custom agents in May of 2024, we’ve seen teams use Rovo to build agents in Confluence that help them accomplish everything from turning customer feedback into PRDs to maintaining consistency across large sets of data and processes.

And agents are getting even more popular, with over 5M invocations of agents every month. They take the foundation of knowledge that lives in Confluence, and act on it, saving customers over 200K hours in February alone. That’s serious time back to help teams ship faster, gain context more easily, and achieve better business outcomes.

We polled organizations to see what agents are adding value to their workflows in Confluence, so you can find inspiration for workflows across Eng, Product, HR, Project Management, and more. Here are seven of our favorites:

• HarperCollins’s Meeting-to-Action Companion: Turns messy notes into crisp decisions, documented in Confluence and completed Jira tickets, so nothing gets lost.
• Docusign’s PRD & Spec Author: Takes a rough brief and ships a review-ready PRD with linked Jira work items in minutes.
• Riverty’s Lessons Learned: Mines past work in Confluence so your team stops re‑learning the hard way.
• KFC’s Architecture Review: Pre-checks proposals against document standards, so approvals are a breeze.
• Pythian’s Progress Tracker: Pulls the signal from Jira + Confluence and drafts status update emails for you.
• Sprout Social’s Onboarding & Playbook Builder: Answers ~80% of new‑hire questions and spins up role guides for new hires.
• Procore’s Backlog & Discovery Synthesizer: Sifts feedback to surface what truly earns a spot on the roadmap.

The seven agents above work across Confluence and Jira — your knowledge layer and your execution layer. But with MCP (Model Context Protocol), agents can reach beyond Atlassian into external tools. An agent can read knowledge in Confluence, think about it, and then take action in a third-party tool without you having to context-switch, reformat, or copy-paste between tabs. Confluence just launched out-of-the-box partner agents, with Lovable, Replit, and Gamma, to turn knowledge into prototypes, codebases, or visuals, all without the human tax of translating between tools.

And because MCP is an open protocol, this ecosystem keeps growing. Any tool that supports MCP can become the next output surface for your Confluence knowledge.

Make your own agents with Rovo Studio

You don’t need to code anything. If you can describe what you want in plain English, you can build an agent. Here’s how:

1. Open Rovo Studio.
   From any Confluence page, click the app switcher in left section of top nav and open Studio. You’ll land on a canvas where you can create and manage agents.

2. Paste your instructions.
   This is the heart of the agent — a plain-language prompt that tells it what to do, what to read, and how to respond. Every use case below includes ready-to-paste instructions you can drop right in.

3. Choose tools and knowledge.
   Pick which tools the agent can use, like Confluence pages, Jira issues, Slack channels, or connect it to tools through the MCP gallery. Under knowledge, scope it to the spaces or projects it needs.

4. Test and publish.
   Run a few test prompts in the preview pane, tweak the instructions until the output feels right, then publish. Your team can start using it immediately from Rovo chat, Confluence pages, or Jira work items.

That’s it — four steps, no engineering ticket required.

Once you’re up and running, here are some tips to get the most out of your agents:

• Pick your use case.
  Identify one or two workflows where Confluence is already the system of record, like PRDs, meeting notes, or incidents, and start with agents there.
• Document your standards.
  Put your templates and guidelines into Confluence pages so agents have clear patterns to follow.
• Use agents with automation.
  Let Studio create a workflow for you. Pick an outcome and use Studio to build an automation rule and agent to let it run. The “Go further” patterns in this post are good starting points and it’s easy to use Natural Language to build the automation rules for you.
• Pilot with a small team.
  Collect feedback on drafts the agents produce. Tune prompts accordingly.
• Scale and measure impact.
  Track time saved on routine docs, number of Confluence pages kept up to standard, and fewer missed follow-ups after meetings.

Want to copy the examples from Docusign, HarperCollins, and Sprout Social? Get started with the prompts and automation patterns below.

First, paste the prompts into Rovo Studio’s Creation screen. These become your agent instructions. Then, confirm all the necessary tools are listed in the Tools drop down, and add the appropriate Spaces and Documents under the Knowledge drop down.

HarperCollins’s Meeting-to-Action Companion

HarperCollins Publishers uses a meeting-to-action agent to turn messy Confluence notes and call transcripts into structured decision logs with owners, deadlines, and linked Jira issues — automatically. It solves the classic loss of follow-ups between meetings by scanning pages, extracting decisions and action items, and creating the right Jira tickets so work is tracked. Ideal for PMs, team leads, and chiefs of staff who live in recurring meetings, it cuts an hour of manual routing and formatting down to ~15 minutes so they can focus on higher‑impact work.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are a Meeting-to-Action Companion agent that turns messy meeting notes and call transcripts into structured decision logs and tracked work.

SCOPE & SOURCES

• Primary notes from [MEETING NOTES SPACE] in Confluence.
• Meeting pages with call transcripts, bullet notes, or freeform text.
• Decision-log standards at [DECISION-LOG STANDARDS PAGE LINK].

WHEN INVOKED

• Read the full notes/transcript and extract:
  • Decisions (what was decided, by whom, and why).
  • Action items (what needs to happen, by when, and by whom).
• Rewrite the page into a structured decision log with sections: Decision, Rationale, Owner, Due date, Links (to Jira issues and related docs).
• For each action item:
  • Create Jira issues using default project [JIRA PROJECT KEY] and issue type [JIRA ISSUE TYPE], unless otherwise specified.
  • Assign issues in this order of precedence: 1) Explicit @mentions on the page, 2) Meeting owner, 3) Team queue or default assignee.
  • Add appropriate labels/components based on the meeting context.
  • Link created Jira issues back to the Confluence page and list their keys under the relevant decision/action.
  • Send a brief Slack notification to [CHANNEL ID] summarizing decisions/actions and linking to the updated page.

OUTPUT

• A cleaned-up Confluence decision log that replaces or augments the original notes.
• A set of Jira issues linked from the decision log.
• A short Slack message with the link and key highlights.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Do not invent decisions, owners, or dates.
• If information is missing, include a short "Open Questions" section.
• Keep sensitive details internal unless explicitly asked for a customer-facing version.

Try these prompts
“Turn these notes into a decision log and create Jira issues for each action.”
“Summarize this customer call into 5 bullets and publish to Confluence.”
“Turn this standup page into a status update for our exec Slack channel.”

Docusign’s PRD & Spec Author

Docusign uses a PRD & Spec Author agent in Confluence to turn a short outline into a complete, review-ready PRD that matches their team’s structure and tone, with suggested Jira epics and stories linked from the doc. It eliminates blank-page PRDs and copy-paste drift by learning from prior specs, decisions, and retros to justify choices. Built for PMs, tech leads, solution architects, and founders who need to go from idea to spec fast, it standardizes outputs, cuts manual rewriting, and reduces glue work across teams.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are a PRD & Spec Author agent that turns short problem statements or Jira work items into complete, review-ready PRDs and specs.

SCOPE & SOURCES

• Prior PRDs and specs in [PRD EXAMPLES SPACE OR LABEL] to learn structure, tone, and depth.
• Standards and templates at [PRD TEMPLATE/STANDARDS PAGE LINK].
• Input brief from a short problem statement and/or a Jira epic (including linked discovery notes and feedback).

WHEN INVOKED

• Analyze the brief and any linked Confluence pages (discovery notes, research, customer feedback).
• Draft a complete, review-ready PRD in Confluence that follows our headings and conventions from [PRD TEMPLATE/STANDARDS PAGE LINK], including:
  • Problem statement and background
  • Goals/non-goals
  • Users and use cases
  • Requirements and acceptance criteria
  • Assumptions
  • Risks
  • Dependencies
  • Open questions
• Propose Jira epics and stories that map to the PRD:
  • Default to project [DEFAULT JIRA PROJECT KEY] and issue types [DEFAULT JIRA ISSUE TYPES], unless otherwise specified.
• Link the Jira issues back to the PRD and cross-link the PRD from the Jira issues.

OUTPUT

• A new or updated PRD in Confluence, published under [WHERE TO PUBLISH PRDS].
• A set of linked Jira epics/stories reflecting the proposed work.
• Optional: a short summary for stakeholders via [NOTIFY VIA] with key highlights and a PRD link.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Stay consistent with tone and level of detail used in prior PRDs in [PRD EXAMPLES SPACE OR LABEL].
• Do not overwrite existing PRDs without preserving key decisions and rationale; update incrementally.

Try these prompts
“Turn these notes into a decision log and create Jira issues for each action.”
“Summarize this customer call into 5 bullets and publish to Confluence.”
“Turn this standup page into a status update for our exec Slack channel.”

Go further: automate it

The most advanced version of this agent doesn’t wait for a PM to invoke it. Set up an automation rule using natural language that triggers on a schedule and points the agent at 90 days of customer feedback from a Jira project or JPD board. The agent identifies emergent themes, clusters them, and auto-generates short PRD drafts for each — structured, user-backed, and ready for a PM to refine.

Automation workflow:
Trigger: Scheduled (weekly) → Action: Invoke “Feedback to PRD & Spec Author” agent → Prompt: “Analyze the last 90 days of feedback from [FEEDBACK PROJECT LINK HERE], identify the top emergent themes, and generate a one-page PRD draft for each theme with linked evidence” → Action: Publish new Confluence page as PRD from agent output.

Riverty’s Lessons Learned Agent

A Lessons Learned Rovo agent helps Riverty turn past investigations into a reusable knowledge base. It scans completed tasks from Riverty’s data analysis desk, distills the key insights, and publishes standardized learnings to Confluence so teams can quickly see what’s been tried before, what worked, and what to avoid. Instead of starting from scratch, users can query the agent with a problem they’re facing and immediately surface lessons gleaned from similar tasks, complete with links back to the original analysis. As Atlassian Product Owner Andrei Tuch puts it, “Every time someone asks us to connect one of the popular LLMs to our Jira or Confluence, we help them implement their use case in Rovo instead – because it always works better.”

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are a Lessons Learned agent that turns completed analysis tasks into a reusable knowledge base and answers new questions by surfacing relevant past learnings.

SCOPE & SOURCES

• Completed tasks from your data analysis desk (Jira tickets, summaries, attachments).
• Confluence pages containing prior analyses, retros, and decision logs.
• Optional: a central "Lessons Learned" index page to organize topics, tags, and links.

WHEN INVOKED

• Ingest and synthesize learnings from completed analysis tasks: problem, approach, data used, key findings, decisions, caveats.
• Publish concise, standardized summaries to Confluence with links back to source tasks and artifacts.
• Answer user queries about current problems by retrieving and summarizing lessons from similar past tasks; highlight applicable insights and known pitfalls.
• When helpful, create or update a central "Lessons Learned" index page that categorizes learnings by topic, system, and tags.

OUTPUT

• A Confluence page (or section) per learning with: Context, What we tried, What worked/failed, Key takeaways, Reuse guidance, Links.
• Cross-links between related learnings and the optional central index.
• Short, actionable answers in chat that cite and link to the relevant learnings.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Preserve source wording for critical details (numbers, thresholds, caveats); summarize without changing meaning.
• Attribute each learning to its original task/page with links and dates.

Try these prompts
“Analyze these tasks and publish Lessons Learned pages in Confluence.”
“Surface past lessons that apply to this problem, with risks and caveats.”
“Update the Lessons Learned index with today’s entries and cross-links.”

KFC’s Architecture Review Agent

KFC uses an Architecture Review agent to raise the quality bar on proposals before they ever reach the Architecture Review Board. The board had strong guidelines and principles in place, but they weren’t consistently followed, so review time was spent sense‑checking completeness and basic alignment instead of debating trade‑offs and long‑term strategy. Now, teams invoke an agent that checks new proposals against KFC’s standards, flags gaps, and suggests improvements, so only proposals that meet baseline expectations move forward. The result: fewer low‑quality submissions, more time on high‑value architectural discussion, and an architecture history that’s easy to navigate and keep up to date.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are an Architecture Review agent that pre-reviews architecture proposals and helps maintain high-quality architecture decision records.

ACCESS

• Confluence (read/write)
• Jira (read-only or write, as configured)

SCOPE & SOURCES

• Architecture principles and guidelines at [ARCHITECTURE PRINCIPLES/GUIDELINES PAGE LINK].
• Example high-quality proposals at [EXAMPLE PROPOSALS SPACE/LABEL/PAGE TREE].
• Architecture decision templates at [ARCHITECTURE DECISION TEMPLATE PAGE LINK].
• Draft proposals as Confluence pages or attachments, optionally linked from Jira.

WHEN INVOKED ON A DRAFT PROPOSAL

• Check the proposal against guidelines, including required sections, clarity of scope and context, explicit trade-offs and alternatives, risks, dependencies, operational concerns, and alignment with existing patterns/principles.
• Compare to similar, high-quality proposals and call out deviations.
• Suggest concrete improvements: sections to add, clarifications to make, risks to articulate, and relevant systems/ADRs/Jira epics to link.
• Provide a short summary at the top with a readiness rating (Ready/Needs Work), key gaps, and recommended next steps.

OPTIONAL: FINAL DECISION RECORD

• Once approved, use [ARCHITECTURE DECISION TEMPLATE PAGE LINK] to generate a finalized Confluence decision record; capture decision, rationale, trade-offs, risks, dependencies; and link related Jira work and ADRs.

OUTPUT

• A reviewed and annotated proposal that meets baseline standards, or clear guidance on what to fix.
• Optionally, a finalized architecture decision record page linked from Jira and other system documentation.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Do not invent architecture decisions, risks, or dependencies; if unclear, add questions for the proposal owner.
• Respect existing decision history; do not overwrite approved decisions, only add context or clarifications.

Try these prompts
“Review this proposal against our guidelines and list what needs fixing.”
“Rewrite this draft to match our architecture proposal template.”
“Create an architecture decision record for this approved proposal.”

Pythian’s Progress Tracker

Pythian uses a Progress Tracker agent to replace ad‑hoc, manual status emails with consistent, data‑backed updates pulled directly from Jira and Confluence. Built for account managers, customer success leaders, and product marketers who send regular customer updates, it scans a Confluence space and linked Jira work to draft polished, customer‑facing summaries with progress, key wins, risks, and next steps — plus an internal‑only version that includes the sensitive details. Instead of hunting across issues, docs, and email threads to remember what changed, teams invoke the agent from a single Confluence page or Jira epic and get tailored updates in minutes. Combined with Transcript Insight and Team Recap agents, it’s saving Pythian teams an average of 20 minutes per day and freeing them up to focus on more strategic, high‑impact work.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are a Progress Tracker agent for customer projects that turns scattered work updates into consistent, data-backed status summaries.

ACCESS

• Confluence (read/write)
• Jira (read for linked projects; write for comments or labels if enabled)

SCOPE & SOURCES

• A specified Confluence page or Jira epic as the system of record for the project.
• Linked Jira issues (status, comments, labels, due dates).
• Related Confluence pages (plans, decision logs, prior updates).
• Example status updates and summaries at [STATUS EXAMPLES SPACE/LABEL/PAGE TREE].

WHEN INVOKED ON A PROJECT

• Determine what has changed since the last update (completed work, in-progress items, blocked tasks, new risks, notable customer interactions/decisions).
• Synthesize a concise status summary emphasizing outcomes delivered, progress against milestones, upcoming work/timelines, and key risks with mitigations.
• When requested, create two versions:
  • Customer-facing: friendly, non-technical, value/timeline/next steps, no sensitive details.
  • Internal: technical details, blockers, staffing notes, technical debt; clearly marked Internal-only.
• Maintain consistent headings: Overview; Progress since last update; Key wins; Risks/blockers; Next steps/upcoming milestones.
• Link relevant Jira issues and Confluence pages in each section.

OUTPUT

• Draft/update a Confluence status page under [STATUS SPACE/PARENT PAGE] with clear date and sections.
• Optional: email/Slack-ready snippets for customer and internal audiences.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Do not expose internal-only details in customer-facing versions.
• Keep tone aligned to prior examples in [STATUS EXAMPLES SPACE/LABEL/PAGE TREE].

Try these prompts
“Draft a customer-ready status update from this plan and linked Jira issues.”
“Summarize progress since the last update with wins, risks, and next steps.”
“Draft an exec summary for the customer and a detailed one for our team.”

Go further: close the loop entirely

The ultimate version of this pattern doesn’t just write the update — it automates the entire feedback collection cycle. Here’s what an internal team at Atlassian built: when new customer feedback hits a Jira project, an automation rule sends the customer a Calendly link to schedule a follow-up call. Loom records the meeting. A Rovo agent summarizes the recording and writes polished notes into a Confluence page. Then a second automation broadcasts the summary — customer name, key themes, action items, and the Loom recording link — to a shared Slack channel. The PM’s only job? Show up and listen. Everything else happens automatically.

Sprout Social’s Onboarding & Playbook Builder

Sprout Social uses an Onboarding & Playbook Builder agent to turn scattered how‑tos, runbooks, and team docs into structured, role‑specific onboarding guides and reusable playbooks. Paired with Loom-based onboarding, the agent now answers ~80% of new‑hire questions directly in their Slack onboarding channel by routing them to the right Confluence content and steps. The result: more than 360 hours of work saved per year, faster ramp-up for new employees, and a dramatically lighter IT and team‑lead burden during those critical first weeks.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are an Onboarding & Playbook Builder agent for new hires and recurring team processes.

ACCESS

• Confluence (read/write)
• Jira (read/write where enabled)
• Slack (read/write in specified onboarding channels)

SCOPE & SOURCES

• Treat the specified Confluence space(s) or page tree(s) as the source of truth for how the team works: how-tos, runbooks, team charters, architecture overviews, decision logs, process docs.
• Use role descriptions, competency frameworks, existing onboarding checklists as anchors for what complete onboarding/playbooks should cover.
• Incorporate Loom or other embedded videos in Confluence as primary learning assets when available.

WHEN INVOKED FOR A NEW HIRE

• Ask for or infer role, team, location, seniority (e.g., "Backend Engineer, Core Services, Mid-level").
• Cluster related Confluence pages into themes (Day 1 basics; Tools & access; Architecture overview; Team rituals; Key projects & metrics).
• Identify gaps or stale content; flag as TODOs with concrete suggestions.
• Draft a sequenced onboarding guide in Confluence (e.g., 1–2 weeks) that states goals, orders topics by need, links canonical docs/Loom/Jira, and proposes tasks that can become Jira issues.

WHEN INVOKED FOR A RECURRING PROCESS PLAYBOOK

• Analyze past Confluence pages, Jira issues, retros for the process (e.g., launches, incidents, QBRs).
• Extract steps, roles, timelines, checklists, known pitfalls.
• Draft a reusable playbook that defines purpose/owner/success metrics; outlines phases with inputs/outputs and DRIs; links examples/templates/Jira; includes a copyable checklist.

SLACK + Q&A BEHAVIOR

• When mentioned in onboarding channels: search relevant Confluence pages and Loom links; point to the best canonical resource first, then summarize if needed.
• If no good answer exists: say so; offer a provisional answer based on adjacent docs; suggest creating/updating a Confluence page/section.

OUTPUT

• A structured onboarding guide page for the role/team.
• One or more reusable process playbook pages in Confluence.
• Helpful, link-rich Slack answers guiding new hires to canonical docs.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Do not invent policies, access rights, or sensitive details; if unclear, add a "Questions for manager or IT" section.
• Prefer existing canonical pages over duplicates; if multiples exist, choose and label the canonical one.

Try these prompts
“Build a 2-week onboarding path for a new backend engineer on this team.”
“Create a reusable playbook for beta launches from these three past docs.”
“Curate the top 10 docs a new PM should read and organize them on one page.”

Procore’s Backlog & Discovery Synthesizer

Procore uses a Backlog & Discovery Synthesizer agent to bridge the gap between customer insights in Confluence and the product backlog in Jira. Instead of PMs trying to remember which interview, feedback log, or research report justified a given idea, the agent connects discovery notes, research pages, and feedback docs to related Jira epics and stories, then rolls everything up into evidence‑backed themes and prioritized recommendations. Prioritization sessions shift from “what do we remember” to “what does the evidence say,” and hunting for buried tickets or docs that used to take 20 minutes now happens almost instantly via Rovo chat and summarized Confluence pages.

Setup instructions

Paste the following into Rovo Studio’s Creation screen.

ROLE

You are a Backlog & Discovery Synthesizer agent that connects customer insights in Confluence to the product backlog in Jira and produces evidence-backed priorities.

ACCESS

• Confluence (read/write)
• Jira (read for relevant projects; write where enabled for comments/labels)

SCOPE & SOURCES

• Discovery and research content in Confluence (user interviews, discovery notes, feedback logs, research reports, experiment results).
• Backlog items in Jira (ideas, feature requests, bugs, roadmap epics) in specified projects/boards.
• Prioritization framework at [PRIORITIZATION FRAMEWORK PAGE LINK] (e.g., RICE, impact/effort).

WHEN INVOKED

• Ask for or infer product area, segment/persona, and timeframe (e.g., "Core workflows, mid-market, last 90 days of feedback").
• Cluster discovery inputs into themes (workflow, persona, segment, problem area).
• For each theme:
  • Identify/link related Jira issues (epics, stories, bugs) via titles, descriptions, tags, linked Confluence pages.
  • Summarize customer signals (pain points, requests, positive feedback) with short excerpts where useful.
  • Flag conflicting/unclear signals as open questions.

OUTPUT

• Create/update a Confluence summary page under [DISCOVERY/ROADMAP SPACE] that lists themes with evidence, groups/links Jira items under each, includes a ranked "Recommended priorities" section with rationale, and captures open questions/risks/assumptions.
• Keep language clear and decision-ready for prioritization and roadmap reviews.

CONSTRAINTS

• Do not invent facts; mark unknowns as TBD and ask focused questions.
• Do not fabricate quotes, quantitative metrics, or prioritization scores; base only on visible data.
• Preserve nuance; if evidence is weak or mixed, state it and recommend further discovery.

Try these prompts
“Synthesize these discovery pages into themes and map them to Jira epics.”
“Recommend top 10 backlog items for next quarter using our RICE framework.”
“Summarize top customer pain points by segment and link related Jira tickets.”

Go further: automate it

One internal team at Atlassian runs a version of this agent on a weekly schedule across 12,000+ feedback tickets spanning five Jira projects and a Slack channel. Every Friday morning, the agent analyzes the last 30 days of feedback, breaks it into weighted themes — positive feedback on new features, negative feedback on usability, feature requests and suggestions — and posts a structured summary to Slack with example quotes, percentages, and specific recommendations for next steps. Leadership reviews it in minutes. No analyst assembled it. No one remembered to ask.

Automation rule:
Trigger: Scheduled (every Friday at 8AM) → Action: Invoke “Feedback Analyst” agent → Prompt: “Analyze the last 30 days of feedback from this project, break it into top themes with examples and links to issues, as well as key customer quotes. Provide recommendations for next steps” → Action: Send Slack message with agent response to your team channel.

How to invoke Confluence Agents with Jira

If your team lives primarily in Jira, these agents meet you there. Confluence is the knowledge layer, Jira is the orchestration layer, and agents move fluidly between the two.

1. Invoke via the assignee picker.
   When you assign an issue in Jira, choose an agent as the assignee. The agent treats the issue as its brief, looks at linked Confluence pages and related tickets, and then writes back into Confluence — creating or updating a PRD, meeting notes page, or whatever the agent is designed to produce — while updating the Jira issue with a link.

2. @mention agents in comments.
   In Jira or (coming soon!) Confluence comments, mention an agent and give it instructions tied to that issue. The agent will read the issue description, attachments, and any linked Confluence pages, perform the requested work, and then respond in the Jira comment thread with links to the Confluence content it created or updated.

3. Use “Open in chat” for deeper refinement.
   From a Jira issue, switch into a chat-style view with the agent using “Open in chat.” This pulls in the issue context and any connected Confluence pages, letting you iterate: ask follow-up questions, refine drafts, or request different versions of a Confluence page or update.

4. Automate invocation with board column triggers.
   Agents can run automatically when an issue moves into certain columns on a Jira board. Moving an epic into “Discovery” could trigger the PRD & Spec Author to draft a first-pass PRD in Confluence. Dragging a bug into “Ready for Postmortem” could trigger the Incident & Postmortem Coach to assemble a Confluence report from linked issues and notes.

5. Enable agents in Jira via Studio surfaces.
   To make agents available inside Jira, you control their visibility through Studio surfaces. By turning on the Jira surface for a given agent, you allow it to appear in the assignee picker, comments, and relevant Jira entry points while keeping its core logic grounded in Confluence and your other knowledge sources.

Original source

The Atlassian Confluence team is pleased to announce the release of Confluence 9.2.20, which is a bug-fix release.

Don't have Confluence 9.2 yet?

Check out the new features and other highlights in the Confluence 9.2 release notes.

Get the latest version

We recommend you read the Confluence 9.2 release notes and you back up your confluence-home directory and database before upgrading.

Key issues fixed in this release include:

• CONFSERVER-103611: Confluence JMX metrics collection (DCDP) logs frequent AttributeNotFoundException: No such attribute: Mean
• CONFSERVER-102539: Confluence uses high L2 cache and OOMs (typically encountered where pages with lots of versions come into play) due to Activity streams gadgets request from Jira
• CONFSERVER-74857: Content REST API is not rate limited

Last modified on May 5, 2026

Original source

The Atlassian Confluence team is pleased to announce the release of Confluence 10.2.11, which is a bug-fix release.

Don't have Confluence 10.2 yet?

Check out the new features and other highlights in the Confluence 10.2 release notes.

Get the latest version

We recommend you read the Confluence 10.2 release notes and you back up your confluence-home directory and database before upgrading.

Issues resolved in 10.2.11

• CONFSERVER-103634: BlueprintDiscoveryUpgradeTask blocks startup with 100% CPU busy (Bug, Highest priority, Fixed)
• CONFSERVER-103611: Confluence JMX metrics collection (DCDP) logs frequent AttributeNotFoundException: No such attribute: Mean (Bug, Medium priority, Fixed)
• CONFSERVER-102709: Scheduled mails gets stuck in Mail Error Queue with OAuth 2.0 SMTP setup (Bug, Low priority, Fixed)
• CONFSERVER-102539: Confluence uses high L2 cache and OOMs due to Activity streams gadgets request from Jira (Bug, Low priority, Fixed)
• CONFSERVER-74857: Content REST API is not rate limited (Bug, Low priority, Fixed)

Last modified on May 5, 2026

Original source

Introducing Remix with Rovo and partner agents in Confluence — a new way to instantly transform Confluence pages into charts, prototypes, presentations, and apps

The last mile of knowledge

Something strange happened over the past decade of work. Teams got incredibly good at creating knowledge — documenting decisions, capturing meeting notes, writing specs. But all that effort exposed a different problem: most of that knowledge never reaches the people who need it, in a format they can actually use.

Confluence pages with visual elements are nearly 2x as likely to be read by a wider audience compared to pages without.

This isn’t a search problem. It isn’t an access problem. It’s a format problem. The knowledge exists. It’s just stuck in a form that doesn’t match how the next person needs to consume it. This means manual work: copying from docs into slides, reformatting for different audiences, and losing context, repackaging existing knowledge instead of creating new.

We’re introducing two new experiences to close that gap to change how teams get value from work they’ve already created. Remix with Rovo transforms content on any Confluence page into new formats like charts, infographics, and other visuals. Pre-built third party partner agents for Lovable, Replit, and Gamma, turn Confluence content into working prototypes, starter apps, and presentations in those tools without manual copy-pasting or custom integrations.

What’s available today

Remix with Rovo begins rolling out today in open beta to Confluence Cloud customers with Rovo, continuing over the next few weeks. At launch, it supports data visualizations, infographics, diagrams, and charts — with more formats coming soon. Find Remix in your Editor toolbar.

Out-of-the-box partner agents for Lovable, Replit, and Gamma are in open beta and start rolling out next week. Admins can enable partner agents in Atlassian Administration under Connected Apps, with no custom agent creation or scripting required.

Introducing Remix with Rovo

Confluence has always been where teams go to create and share knowledge. With Remix, it becomes something more: an adaptive workspace — one where the content itself reshapes to meet the reader, not the other way around.

Select any content on a Confluence page and instantly transform it into a visual format optimized for how someone needs to consume it.

A data-heavy section becomes a chart. A process description becomes an infographic. A long-form analysis becomes a visual summary. No copy-pasting, no switching tools, no reformatting. Just the boost in understanding that comes from nailing the format.

Confluence pages with 1 or more visual element are 18% more likely to be read by a wider audience.

Three things make Remix with Rovo fundamentally different from what’s come before:

1. It’s non-destructive. Remix never overwrites your page. Every remix is an extra layer on top of the source, which stays intact as the canonical version — so you get new ways to view the content without creating copies that go stale.

2. It’s opinionated. Remix gives you ready-made format options or a freeform prompt if you already know what you want. Pick a preset (like a chart for numbers or an infographic for a flow), or describe the output in your own words. In both cases, it analyzes the content to propose a strong first version you can tweak.

3. It’s embedded, not separate. Remix views are created and live right on the page. Instead of sending people to a separate deck, report, or tool, the most digestible version of the content sits where they already are. Anyone visiting a Confluence page can turn the source into the version that’s easiest for them to scan, compare, and act on.

When the right format lives in a different tool

Sometimes the next step isn’t a better chart — it’s a working prototype. A starter app.

That’s why we’re also launching out-of-the-box partner agents in Confluence, starting with Lovable, Replit, and Gamma, built on Rovo and powered by MCP.

From any Confluence page, invoke a partner agent that carries your content (and context) into a native output in that partner’s tool using Rovo Chat. A product spec becomes a real Lovable application our designer can interact with in minutes. A technical doc becomes a Replit starter app your engineer can fork and extend. Meeting notes become a Gamma presentation your team lead can walk into a room with.

And Rovo Skills keep that output linked back to the source page it came from. That linkage runs through the Teamwork Graph, the same layer of work relationships and context, built from over 100 billion data points across Atlassian, that powers agents in Jira and MCP skills for Rovo. When a partner agent carries your content into Lovable or Replit, it doesn’t just carry the text. It carries the context: who created it, what project it belongs to, what decisions it connects to.

Enable a partner’s MCP server once and within minutes, teams get a ready-to-use agent in their Rovo directory, pre-configured by the partner, inheriting the permissions and context of your workspace. And because everything routes back through Confluence, work created in an external tool doesn’t disappear into that tool’s silo. It stays anchored to your source of truth.

And these partner agents are just the beginning.

Go further with MCP skills in Rovo

The same foundation that’s powering these Partner agents – MCP – also lets you bring in tools beyond the ones we’ve launched with today.

MCP lets any tool connect to Confluence as an AI-aware service. Today that includes Lovable, Replit, and Gamma. But the protocol is open, the server is documented, and any partner can build an agent that works with the knowledge your team already has in Confluence, without waiting for us to build a bespoke integration.

To discover MCP‑compatible skills from your favorite apps, and use them with Rovo across Confluence, Jira, and more, visit our gallery of MCP servers and start connecting them to your work today.

Check out Rovo MCP skills

A different bet about AI in the enterprise

This is the second chapter of a platform shift we started in February. Agents in Jira showed what happens when AI joins your team inside the tool where work gets tracked. Today, Remix and partner agents show what happens when AI joins your team inside the tool where knowledge lives. Together, they mark a turn from AI that helps individuals produce faster to AI that helps teams deliver to each other — across tools, across formats, across the last mile.

Because the last mile of knowledge isn’t about writing more. It’s about delivering better.

See what that looks like in practice

Check out our new digital series, Rovo at Work, to see product demos and real-world examples of how Atlassian teams use Remix, Rovo Skills, and Rovo Dev in Jira to transform how they get work done and deliver better outcomes.

Watch Rovo at Work

Original source

Introducing Remix with Rovo and partner agents in Confluence — a new way to instantly transform Confluence pages into charts, prototypes, presentations, and apps

The last mile of knowledge

Something strange happened over the past decade of work. Teams got incredibly good at creating knowledge — documenting decisions, capturing meeting notes, writing specs. But all that effort exposed a different problem: most of that knowledge never reaches the people who need it, in a format they can actually use.

Confluence pages with 1 or more visual element are 18% more likely to be read by a wider audience.

This isn’t a search problem. It isn’t an access problem. It’s a format problem. The knowledge exists. It’s just stuck in a form that doesn’t match how the next person needs to consume it. This means manual work: copying from docs into slides, reformatting for different audiences, and losing context, repackaging existing knowledge instead of creating new.

We’re introducing two new experiences to close that gap to change how teams get value from work they’ve already created.

Remix with Rovo transforms content on any Confluence page into new formats like charts, infographics, and other visuals.

Pre-built third party partner agents for Lovable, Replit, and Gamma, turn Confluence content into working prototypes, starter apps, and presentations in those tools without manual copy-pasting or custom integrations.

What’s available today

Remix with Rovo starts rolling out today in open beta for Confluence Cloud customers with Rovo. At launch, Remix supports — data visualizations, infographics, diagrams, and charts — with more formats coming soon.

Out-of-the-box partner agents for Lovable, Replit, and Gamma are in open beta and start rolling out next week. Admins can enable partner agents in Atlassian Administration under Connected Apps, with no custom agent creation or scripting required.

Introducing Remix with Rovo

Confluence has always been where teams go to create and share knowledge. With Remix, it becomes something more: an adaptive workspace — one where the content itself reshapes to meet the reader, not the other way around.

Select any content on a Confluence page and instantly transform it into a visual format optimized for how someone needs to consume it.

A data-heavy section becomes a chart. A process description becomes an infographic. A long-form analysis becomes a visual summary. No copy-pasting, no switching tools, no reformatting. Just the boost in understanding that comes from nailing the format.

Confluence pages with 1 or more visual element are 18% more likely to be read by a wider audience.

Three things make Remix with Rovo fundamentally different from what’s come before:

1. It’s non-destructive. Remix never overwrites your page. Every remix is an extra layer on top of the source, which stays intact as the canonical version — so you get new ways to view the content without creating copies that go stale.

2. It’s opinionated. Remix gives you ready-made format options or a freeform prompt if you already know what you want. Pick a preset (like a chart for numbers or an infographic for a flow), or describe the output in your own words. In both cases, it analyzes the content to propose a strong first version you can tweak.

3. It’s embedded, not separate. Remix views are created and live right on the page. Instead of sending people to a separate deck, report, or tool, the most digestible version of the content sits where they already are. Anyone visiting a Confluence page can turn the source into the version that’s easiest for them to scan, compare, and act on.

When the right format lives in a different tool

Sometimes the next step isn’t a better chart — it’s a working prototype. A starter app.

That’s why we’re also launching out-of-the-box partner agents in Confluence, starting with Lovable, Replit, and Gamma, built on Rovo and powered by MCP.

From any Confluence page, invoke a partner agent that carries your content (and context) into a native output in that partner’s tool using Rovo Chat. A product spec becomes a real Lovable application our designer can interact with in minutes. A technical doc becomes a Replit starter app your engineer can fork and extend. Meeting notes become a Gamma presentation your team lead can walk into a room with.

And Rovo Skills keep that output linked back to the source page it came from. That linkage runs through the Teamwork Graph, the same layer of work relationships and context, built from over 100 billion data points across Atlassian, that powers agents in Jira and MCP skills for Rovo. When a partner agent carries your content into Lovable or Replit, it doesn’t just carry the text. It carries the context: who created it, what project it belongs to, what decisions it connects to.

Enable a partner’s MCP server once and within minutes, teams get a ready-to-use agent in their Rovo directory, pre-configured by the partner, inheriting the permissions and context of your workspace. And because everything routes back through Confluence, work created in an external tool doesn’t disappear into that tool’s silo. It stays anchored to your source of truth.

And these partner agents are just the beginning.

Go further with MCP skills in Rovo

The same foundation that’s powering these Partner agents – MCP – also lets you bring in tools beyond the ones we’ve launched with today.

MCP lets any tool connect to Confluence as an AI-aware service. Today that includes Lovable, Replit, and Gamma. But the protocol is open, the server is documented, and any partner can build an agent that works with the knowledge your team already has in Confluence, without waiting for us to build a bespoke integration.

To discover MCP‑compatible skills from your favorite apps, and use them with Rovo across Confluence, Jira, and more, visit our gallery of MCP servers and start connecting them to your work today.

Check out Rovo MCP skills

A different bet about AI in the enterprise

This is the second chapter of a platform shift we started in February. Agents in Jira showed what happens when AI joins your team inside the tool where work gets tracked. Today, Remix and partner agents show what happens when AI joins your team inside the tool where knowledge lives. Together, they mark a turn from AI that helps individuals produce faster to AI that helps teams deliver to each other — across tools, across formats, across the last mile.

Because the last mile of knowledge isn’t about writing more. It’s about delivering better.

See what that looks like in practice

Check out our new digital series, Rovo at Work, to see product demos and real-world examples of how Atlassian teams use Remix, Rovo Skills, and Rovo Dev in Jira to transform how they get work done and deliver better outcomes.

Watch Rovo at Work

Original source

The Atlassian Confluence team is pleased to announce the release of Confluence 10.2.8, which is a bug-fix release.

Don't have Confluence 10.2 yet?

Check out the new features and other highlights in the Confluence 10.2 release notes.

Get the latest version

We recommend you read the Confluence 10.2 release notes and you back up your confluence-home directory and database before upgrading.

Issues resolved in 10.2.8

• CONFSERVER-102561: Pages created using template storage format via the REST API appear blank in the editor (Bug, Medium priority, Fixed)
• CONFSERVER-102546: "Manage Watchers" button doesn't work and throw console exceptions (Bug, Low priority, Fixed)
• CONFSERVER-101960: In the Jira Issue/Filter macro, Target Start and End dates are behind by 1 day when server is set to UTC and Customer Timezone is GMT-8 (Bug, Low priority, Fixed)
• CONFSERVER-101847: Confluence cluster abruptly loses the ability to connect to Confluence database via Kerberos authentication (Bug, Low priority, Fixed)
• CONFSERVER-100549: Space query when using OpenSearch as the search platform, requests all stored fields (Bug, Low priority, Fixed)
• CONFSERVER-100318: Sandbox processes are not terminated by Confluence if they can't be gracefully shutdown within 20 seconds (Bug, Low priority, Fixed)
• CONFSERVER-99593: The completion date for the Task report macro is not captured under the Profile task report (Bug, Low priority, Fixed)
• CONFSERVER-56438: Modifying Additional User Details on User Profile Page Triggers Application Permission Validation (Bug, Low priority, Fixed)

Original source

In the run‑up to Team ’25 Europe, I was working on how to create content with Rovo, an AI‑powered creation experience that turns prompts plus context into ready‑to‑use work in Confluence: pages, plans, PRDs, briefs, and more.

As a Product Marketer, my job wasn’t just to make it look magical in a demo. It was to make sure it actually helped real people do their jobs better.

For Team ’25 Europe, we wanted creating with Rovo at the center of our story. That gave us two clear objectives:

1. The product needed to work reliably in realistic conditions
2. It needed to showcase clear, undeniable value on stage

Essentially, we couldn’t fake it with cherry‑picked moments. The experience had to hold up in the same messy environments where teams actually work, with real stakes and real expectations.

To get there, I partnered closely with Product, Engineering, and other Marketers to optimize the create with Rovo experience so it produced high‑quality outputs tailored to realistic, specific use cases by persona. Together, we built a set of “golden prompts” and story‑first demo scenarios across Marketing, Sales, Program Management, Product, and Engineering, which our Engineers could then rigorously pressure‑test — until we trusted them live.

In this post, I’ll walk through:

• What “golden prompts” are — and when they actually matter
• How to design prompts around real workflows and personas (with examples)
• A repeatable process for testing AI outputs across tools
• How we mapped realistic demo scenarios using Confluence Whiteboards
• How cross‑functional teams share responsibility for AI‑assisted decisions

What is a “golden prompt” — and when should you use one?

A golden prompt is a specifically phrased request that reliably produces a high‑quality, high‑stakes output for a well‑defined use case.

We knew we didn’t need unrealistically clever prompts. We needed the right prompts — the ones that would matter when people were actually accountable for outcomes. It means:

• The stakes are real (the output will be used in front of customers, leaders, or in live execution).
• The goal is clear (we can describe what a “good” output looks like in advance).
• The inputs are grounded in trustworthy context (work data, docs, and examples).
• The output is repeatable across tools and runs, not a one‑off lucky win.

Golden prompts are not for every AI interaction. If the work is one‑off, low‑risk, or easily fixed by a single person, you probably don’t need golden prompts yet. As soon as AI starts influencing shared plans, decisions, or customer‑visible work, it’s worth treating a small set of prompts as strategic assets.

Why we started with golden prompts (with examples)

In other words: golden prompts aren’t about showing off what AI can do in a vacuum. They’re about designing the few, critical interactions where AI touches decisions, customers, and delivery in a visible way.

We defined 10–15 golden prompts per persona. Each one captured:

• A specific outcome (campaign plan, PRD, sprint board, report)
• The real‑world constraints (channels, segments, timeframes, stakeholders)
• The minimum quality bar we’d be proud to show live and ship to real teams

Here are a few golden prompt examples.

For Marketing

• Create a reactivation campaign plan for [user segment], using [marketing channels], referencing their previous usage and new features released since they left
• Create a webinar outline for [topic] targeting [audience]. Include: compelling title, 5-slide agenda, key talking points for each section, audience engagement moments (polls/Q&A), strong CTA, follow-up resources.

For Sales

• Create a report extracting from [CRM tool] to include my deals for [timeframe], including the customer, segment, ACV/TCV, and a summary of each
• Analyze previous call recordings, footprint, and CRM notes on [customer] to further personalize discovery questions tailored to this account towards a potential solution

For Product

• Create a PRD for [feature]. Include sections for 1. User Stories (at least 3, in ‘As a… I want… so that…’ format), 2. Acceptance Criteria for each story, 3. Success Metrics (KPIs) for the feature, and 4. Technical Considerations
• Identify 3-5 emerging trends relevant to [theme/category]. For each trend, explain its potential impact on our product strategy and suggest a strategic response

For Engineering

• Create a sprint planning whiteboard for the upcoming release cycle for [project/launch], listing team members, sprint goals, prioritized backlog items, story point estimates, and capacity planning
• Generate API documentation for [Service]. Include details for Request Parameters, Response Body (JSON schema), Error Codes (404, 500), and a cURL example.

For any Knowledge Worker

• Start a Retro Board for the upcoming retrospective session with my team
• Create an Out of Office page for these [dates], with the right back ups for the priorities I’m working on

We treated these golden prompts as an alignment device. They forced clarity on:

• The user’s actual goal
• The available context (first‑ and third‑party knowledge, work data)
• What “great” looks like in the final artifact — not just “good enough” for a screenshot

These prompts became our north star for quality. If creating content with Rovo could reliably handle these, we knew we were building something teams could trust when the stakes were real.

How we defined prompt “quality” — and why it mattered

Next, a braintrust of folks across Marketing, Product, and Engineering tested the prompt output relentlessly across multiple AI tools, including ChatGPT, Claude, Gemini, Perplexity, and our very own Rovo. We documented outputs in Confluence and tracked progress in Jira. Using multiple tools served a purpose:

• It reduced over‑reliance on any single model
• It helped us spot recurring weaknesses (e.g., missing sections, hallucinated details)
• It showed where our context and prompts needed tightening

The bar was simple and unforgiving: if a golden prompt didn’t yield an output we’d be proud to show live—and put in front of our customers!—we refined the prompt, clarified the task, or improved the context. We also merged outputs from various AI tools to get the best possible outcome.

For us, a “high‑quality” output needed to:

1. Work consistently across tools and runs
2. Match the real needs of that persona (not an abstract persona sketch)
3. Be immediately actionable in the work context where it would be used

Our step‑by‑step Golden Prompt evaluation process

We ended up with a repeatable process that any team can adapt:

1. Define target output and audience upfront
   Who will use this, and what decision or task should this output move forward?
2. Specify inputs clearly
   What context should AI pull from? What’s in scope, out of scope, or off‑limits?
3. Run across multiple tools
   Compare outputs for clarity, structure, and actionability. Note where each tool struggles or overreaches.
4. Tighten phrasing to remove ambiguity
   Replace soft asks (“help with”, “improve”) with specific instructions, formats, and required sections.
5. Re‑test until the output is top‑tier
   If the output still feels like something you’d hesitate to share with stakeholders, keep iterating.

This process did more than improve prompts. It forced shared ownership of what “good” looks like — and where AI should stop so humans can take over.

Designing real‑world scenarios with Confluence whiteboards

Here’s where Product Marketing could really shine. I mapped our top scenarios end‑to‑end to bring the creating with Rovo experience to life:

• Where the human starts (an executive ask, a noisy board, a half‑baked idea)
• How Rovo shapes the work (synthesis, structure, first drafts, options)
• Where alignment happens (review, critique, decisions)
• How outcomes turn into artifacts (plans, PRDs, boards) that drive execution

We focused on two persona stories in particular:

• Marketing: turning executive asks into collaborative outreach plans and content
• Product: synthesizing research, generating PRDs, and orchestrating work across teams

In both journeys, we deliberately toggled between:

• Divergent phases (brainstorming, exploration, synthesis)
• Convergent phases (formatting, structuring, decisions, and next steps)

That produced plausible demos and “magic moments.” This is where many “AI stories” break down in the real world—showing the messy back‑and‑forth prompting where humans and AI refine, question, and adjust together.

Once our prompts were more stable, we optimized scenarios for the live event environment. Using Confluence whiteboards first before creating screens and animations helped us define human/AI handoffs, decide where to pause for impact and where to streamline so the story kept momentum.

By mapping those loops on a whiteboard first — before we built screens, animations, or keynote flows — we could:

• Make human/AI handoffs explicit
• Decide where to pause for impact in the live demo
• Streamline so the story kept its momentum

The result was demos that felt real because they were real: grounded in workflows that would still make sense the morning after the event.

Why this work went beyond “a good demo”

From the outside, this might look like “just” a demo prep story. In reality, the work reflects how we think about AI‑assisted experiences more broadly.

• Golden prompts are most useful where accountability is highest.
  These aren’t everyday experiments. They’re the prompts that produce plans, reports, and decisions you’d still stand behind weeks later.
• Quality is a shared responsibility, not a prompt‑writer’s problem.
  Our strongest prompts came from combining product truth, technical constraints, and go‑to‑market realities — not from a lone expert tweaking wording in isolation.
• Scenario design matters as much as model choice.
  By mapping journeys end‑to‑end, we saw exactly where AI should help, where it should defer, and where humans must remain clearly responsible.

Cross‑functional partnership: our unfair advantage

The best prompts didn’t come from one person. They came from a cross‑functional braintrust that stayed involved from idea to stage to product:

• Product Managers brought the product truth: what’s feasible today, what’s coming, and which workflows actually matter to teams.
• Engineers stress‑tested feasibility and edge cases, and helped us understand how prompts interacted with real data and systems.
• Product Marketing kept us anchored on real use cases, narrative flow, and user relevance, making sure we were solving for accountable, not hypothetical, work.

Using our cross-functional braintrust early paid off immediately. It showed up in:

• The confidence we had going into the live demo — we weren’t hoping it would work, we’d already done the hard testing
• The shared standard we built around responsible, reliable AI in the places where delivery actually has consequences.
• The experience customers get now, when they create content with Rovo in the same kinds of scenarios we rehearsed on stage

The good news is that creating and editing with Rovo is now live and available to all! Read this blog to learn more, and start creating content with Rovo today by clicking here.

This work was deeply cross‑functional, and I owe a huge thank‑you to everyone who helped us design, test, and harden these golden prompts.

Product management: Aniket Vaidya (who led much of the early work) and John Murnen
Engineering: Paul Borza, Velu Alagianambi, Peter Martigny, and Dhanraj Jadhav
Sales: Scott Silver
Marketing: Shelley Wang and Nidhi Chaudhry
And our Program Managers: Manjiri Soman and Guru Prakash Nagarajan

Original source

The Atlassian Jira team is pleased to announce the release of Jira Software 11.3.3.

Don't have Jira Software 11.3.x yet?

Check out the new features and other highlights in the Jira Software 11.3.x release notes.

Get the latest version

Control who has access to automation components

Starting from Jira 11.3.3, you can use automation restrictions to decide who can create, edit, enable, or disable automation rules that use specific components. This feature gives you more control over automation and is useful when project admins manage project rules. Only users in specified groups can manage rules with restricted components.

How to manage automation restrictions

Released on 5 March 2026

16 issues

Last modified on Mar 5, 2026

Original source

The Atlassian Confluence team is pleased to announce the release of Confluence 9.2.17, which is a bug-fix release.

Don't have Confluence 9.2 yet?

Check out the new features and other highlights in the Confluence 9.2 release notes.

Get the latest version

We recommend you read the Confluence 9.2 release notes and you back up your confluence-home directory and database before upgrading.

Key issues fixed in this release include:

• CONFSERVER-102567: Injection dompurify Dependency in Confluence Data Center (Public Security Vulnerability, High priority, Fixed)
• CONFSERVER-101133: viewinfo.action will trigger expensive permissions check for pages with lots of incoming links (Bug, Medium priority, Fixed)
• CONFSERVER-99603: Editing Mode Automatically Scrolls Up When Using a Firefox or Chrome Browser (Bug, Medium priority, Fixed)
• CONFSERVER-96206: Referencing a sufficiently large PowerPoint/ Word/ Xls/ Pdf file on a Confluence Data Center page through the Office connector macro can cause the instance to run out of memory (Bug, Medium priority, Fixed)
• CONFSERVER-94493: StackOverflowError when exporting large pages to PDF (Bug, Medium priority, Fixed)
• CONFSERVER-101456: Date macro allows the insertion of a 5+ digit year-number (like 20215) via the "//" shortcut, prevents content from being saved, and throws a DateTimeParseException in the app logs (Bug, Low priority, Fixed)
• CONFSERVER-100857: Space Export to PDF function in Confluence, where links in the exported PDF no longer navigate within the document but instead open web links (Bug, Low priority, Fixed)

Last modified on Mar 3, 2026

Original source

Release notes for Confluence 10.2.7

The Atlassian Confluence team is pleased to announce the release of Confluence 10.2.7, which is a bug-fix release.

Don't have Confluence 10.2 yet?

Check out the new features and other highlights in the Confluence 10.2 release notes.

Get the latest version

We recommend you read the Confluence 10.2 release notes and you back up your confluence-home directory and database before upgrading.

Released on 03 March 2026

Key Summary T Created Updated Due Assignee Reporter P Status Resolution CONFSERVER-96206 Referencing a sufficiently large PowerPoint/ Word/ Xls/ Pdf file on a Confluence Data Center page through the Office connector macro can cause the instance to run out of memory Bug Jul 19, 2024 14:32 Mar 03, 2026 09:16 Jeffery Xie Basar Beykoz Medium CLOSED Fixed CONFSERVER-94493 StackOverflowError when exporting large pages to PDF Bug Feb 13, 2024 11:23 Mar 03, 2026 09:16 Jing Zheng Ravi Dixit Medium CLOSED Fixed CONFSERVER-102204 Upgrading to Confluence 10.2.x (with Oracle 19c) fails when Faster Permissions was enabled Bug Feb 18, 2026 22:28 Mar 03, 2026 09:16 Saquia Naz Eric L Low CLOSED Fixed CONFSERVER-101456 Date macro allows the insertion of a 5+ digit year-number (like 20215) via the "//" shortcut, prevents content from being saved, and throws a DateTimeParseException in the app logs Bug Nov 06, 2025 01:01 Mar 03, 2026 09:16 Jake Lyell Lei Wang Low CLOSED Fixed CONFSERVER-97548 Email notifications are not sent for personal spaces Bug Aug 06, 2024 07:45 Mar 03, 2026 09:16 Saba Taseer Ozhan Aydar Low CLOSED Fixed

5 issues

Last modified on Mar 3, 2026

Original source

Confluence 10.2.7

The Atlassian Confluence team is pleased to announce the release of Confluence 10.2.7, which is a bug-fix release.

Don't have Confluence 10.2 yet?

Check out the new features and other highlights in the Confluence 10.2 release notes.

Get the latest version.

We recommend you read the Confluence 10.2 release notes and you back up your confluence-home directory and database before upgrading.

Key issues fixed in 10.2.7 include:

• CONFSERVER-96206: Referencing a sufficiently large PowerPoint/ Word/ Xls/ Pdf file on a Confluence Data Center page through the Office connector macro can cause the instance to run out of memory.
• CONFSERVER-94493: StackOverflowError when exporting large pages to PDF.
• CONFSERVER-102204: Upgrading to Confluence 10.2.x (with Oracle 19c) fails when Faster Permissions was enabled.
• CONFSERVER-101456: Date macro allows the insertion of a 5+ digit year-number (like 20215) via the "//" shortcut, prevents content from being saved, and throws a DateTimeParseException in the app logs.
• CONFSERVER-97548: Email notifications are not sent for personal spaces.

All issues are marked as fixed and closed as of March 3, 2026.

Original source

Release notes for Confluence 9.2.17

The Atlassian Confluence team is pleased to announce the release of Confluence 9.2.17, which is a bug-fix release.

Don't have Confluence 9.2 yet?

Check out the new features and other highlights in the Confluence 9.2 release notes.

Get the latest version

We recommend you read the Confluence 9.2 release notes and you back up your confluence-home directory and database before upgrading.

Issues resolved in 9.2.17:

• CONFSERVER-101133: viewinfo.action will trigger expensive permissions check for pages with lots of incoming links (Bug, Medium, Fixed)
• CONFSERVER-99603: Editing Mode Automatically Scrolls Up When Using a Firefox or Chrome Browser (Bug, Medium, Fixed)
• CONFSERVER-96206: Referencing a sufficiently large PowerPoint/ Word/ Xls/ Pdf file on a Confluence Data Center page through the Office connector macro can cause the instance to run out of memory (Bug, Medium, Fixed)
• CONFSERVER-94493: StackOverflowError when exporting large pages to PDF (Bug, Medium, Fixed)
• CONFSERVER-101456: Date macro allows the insertion of a 5+ digit year-number (like 20215) via the "//" shortcut, prevents content from being saved, and throws a DateTimeParseException in the app logs (Bug, Low, Fixed)
• CONFSERVER-100857: Space Export to PDF function in Confluence, where links in the exported PDF no longer navigate within the document but instead open web links (Bug, Low, Fixed)

Original source