Jira tickets for AI: Atlassian lets agents take over

Contents

At its in-house conference Team '26, Atlassian presented several new features designed to integrate AI agents more deeply into users' daily collaboration. The focus is on expanding the Teamwork Graph as an enterprise-wide context layer and further developing the AI platform Rovo. Its agents are intended to not only support tasks in the future but also to plan and execute them independently.

Atlassian develops tools for collaboration and software development. These include Jira, Confluence, and Loom. All these tools are intended to bring together tasks, knowledge, and communication within teams.

Teamwork Graph opens up to external agents

The Teamwork Graph plays a central role. It maps relationships between tasks, documents, people, and systems, providing AI agents with the necessary context. What's new is that this context is no longer available only within Atlassian products. Developers can access the Graph directly in the terminal via a new command-line tool – currently in open beta. Additionally, Atlassian provides interfaces via the Model Context Protocol (MCP), allowing external agents and copilots to use the data. This enables AI systems to incorporate relationships such as responsibilities, dependencies, or previous decisions, rather than answering isolated queries. For example, an agent can determine which incidents are related to a specific deployment and who is responsible for resolving them.

In parallel, Atlassian is expanding Rovo – an AI-powered search and knowledge discovery tool – and evolving it from a pure assistance tool to a tool for agentic work. AI agents are intended to independently break down, plan, and execute complex, multi-stage tasks. The already announced reasoning mode “Max” in Rovo Chat will orchestrate such processes across multiple tools in the future. An example of practical application would be a quarterly report for which an agent compiles data from various sources, processes it, and flags missing information.

AI agents move into existing workflows

AI agents are also moving closer to existing workflows. In Jira, tasks can now be specifically assigned to agents (called Agents in Jira, already generally available), who will process or prepare them independently. In Confluence, the Remix function converts content into other formats such as presentations or diagrams without users having to leave the environment. Loom converts video instructions into structured tasks that can be further processed, for example, as Jira tickets. Point-based AI queries are thus intended to give way to permanently embedded automation.

With Rovo Studio, Atlassian also offers a no-code platform on which users can create their own agents, automations, and applications. It builds on the Teamwork Graph and is explicitly not just for developers. Workflows can be defined based on events and secured with functions such as role models, approvals, and versioning. An example would be an onboarding process where an agent automatically creates accounts, provides documents, and distributes tasks as soon as a new employee is registered in the system.

More transparency for development teams

Atlassian is specifically expanding its offering in the Developer Experience (DX) area for software developers. New features such as “Agent Experience,” “AI Code Insights,” and “AI Pulse” are intended to create transparency regarding AI usage in the development process. This allows tracking which portion of the code originates from AI, how agents are integrated into workflows, and how this affects productivity and quality.

With the Product Collection, Atlassian is also announcing a new product line for product management. It expands existing tools such as Jira Product Discovery and is intended to cover the entire process from collecting customer feedback, prioritization, to implementation and success measurement.

Also new are the Dia Reports, which are browser-based briefings that combine information from the Teamwork Graph with data from typical work tools such as calendars or communication platforms. This results in automatically generated daily summaries that bundle open tasks, relevant discussions, and upcoming appointments.

More details on all new features can be found in the Atlassian blog.

(fo)