Most AI is built for a single person in a single moment. But work is shared, continuous, and cumulative across teams.
What starts as an engineering sprint in Jira evolves into a shared marketing launch plan in Confluence, then lands in Jira Service Management as ready-to-use customer support knowledge, all the while touching multiple connected apps. One thread across three different teams.
Rovo already understands how your work connects across teams, tools, and projects via the Teamwork Graph, and remembers what matters to you.
We are introducing a new set of Rovo Chat capabilities: visible memory controls, @mentionable agents, shareable chats, and custom skills that turn repeated work into reusable workflows.
Today’s new chat features push our team-first foundation even further, helping context carry forward, expertise remain accessible, and work stay connected across the apps teams use.
Memory you can see and shape: control what Rovo remembers
The more your team uses Rovo, the more it can recall for you. Rovo’s memory has always worked in the background, learning your preferences and how you work. Now you can see what it has learned and shape it yourself.
A new memory view shows you exactly what Rovo knows: that you want summaries in bullets, that a certain project shipped in Q1, that customer data stays out of drafts – ensuring you don’t have to repeat yourself or reshare context. You can view and make edits to what Rovo knows, adjust your preferences, and how Rovo responds.
You and your admin can toggle memory on or off at any time at both the user and site level.
@mention agents: bring the right expertise into your conversation
Sometimes the work reaches a point where a specialist needs to take over: a competitive intel agent, a launch planning agent, a data analyst agent. Now you can @mention any agent directly in Rovo Chat, and it gets right to work with the full context of what you’ve already worked through.
Once you bring an agent in, it has the thread, the decisions, and the artifacts it needs to pick up where you left off.
Share chats: hand off the full thinking to your team
You worked out an analysis, a plan, and the background to make a decision. Handing off that work to another teammate usually means writing a summary that loses most of the thinking that got you there.
When one person figures something out, the whole team should be able to build on it, not start from scratch. Now you can share the Chat itself with the relevant context. Your teammate see how you got there and picks up where you left off – no back and forth needed.
Information shared in Chat remains permissions-aware so teammates only see what they already have permission to see and nothing more.
Custom skills for workflows: codify processes
Every team has its own standards, language, and repeatable workflows, like the steps to triage a bug or the right way to summarize weekly updates for leadership. While AI can help with speed and execution, it still needs to know how your team likes to work.
Now you can encode those processes and knowledge as custom skills. Describe what you want in Chat, “create a skill that generates a sprint summary from my board”, and Rovo builds it by asking questions to understand edge cases and how you want the result to look.
Create, refine, and update skills in Chat or in Rovo Studio, all in natural language.
(Custom skills will be available across all instances over the coming week.)
Start somewhere this week
Great work is rarely a solo sport. It takes teams, agents, and shared context to keep momentum going.
@mention an agent when you need specialized expertise on the fly. Share a chat so your team picks up right where you left off. Teach Rovo a skill for a workflow you repeat every week.
Context is only valuable when it’s shared. Try it out in chat today.
