Atlassian launches Jira tools for agentic software teams

Atlassian has launched new Jira and DX features for agentic software engineering workflows, aimed at helping engineering teams manage AI use across the software development lifecycle.

The new functions are intended to give software teams more control over how AI agents access context, carry out tasks, and are assessed within existing development processes. The rollout covers planning, coding, review, and measurement, with features arriving through open beta, private early access, and staged general availability.

At the centre of the launch is Atlassian's effort to make AI agents work from shared organisational context rather than isolated prompts. It is introducing Code Context, built on its Teamwork Graph, to give Rovo and coding agents access to information across multi-repository codebases. Atlassian said this should help agents use architecture, requirements, and project information when assessing backlog items, drafting implementation plans, investigating defects, and tracing root causes.

A second control layer, Agent Context Controls, is intended for platform teams that want to limit which Jira and Confluence spaces agents can use. The goal is to keep outputs aligned with approved requirements, architecture decisions, and internal standards.

Backlog automation

Another part of the launch focuses on moving agentic work from ad hoc use into recurring workflows. Atlassian is adding agent loops in Jira, which it said can scan for well-defined unassigned backlog tasks, pass them to the Jira Coding Agent for execution and testing, and then open pull requests in Jira for review.

It is also adding a feature called Standards, allowing platform teams to define coding standards once and link them to repositories. Atlassian said this creates a common rule set for both human developers and software agents working in the same codebase.

Alongside that, AI review introduces a separate agent to examine pull requests against organisational standards before code is merged. Together, these changes are meant to create a more structured path from task definition to code review, while leaving final approval with developers.

Measurement and governance

Atlassian is also extending its DX product with a package called DX for Agentic Development. It said the product measures AI use across throughput, quality, adoption, and cost, while linking AI spending to engineering outputs. It includes AI Code Insights, tracking for tools and model context protocol systems, analysis of model-to-task fit, and Agent Experience research.

For team leaders using Jira, the new Agent Usage Dashboard is intended to show which agents are appearing in workflows and how those sessions relate to Jira work items. Atlassian said the dashboard is designed to help managers connect agent activity with changes in delivery performance.

The move reflects a broader challenge for software teams that have begun using AI in development but have struggled to expand that use in a controlled way. Atlassian cited findings from its 2026 AI SDLC study showing that 94% of engineering leaders report using AI, while only 6% have systems in place to scale it across the full software development lifecycle.

That gap has become more pressing as organisations move from one-off code generation and chatbot sessions to persistent workflows in which agents can plan, write, test, and review work with limited supervision. In that setting, companies need ways to decide what agents can access, what tasks they can perform, and how their output should be checked.

Atlassian also pointed to DX analysis showing that teams whose AI tools used the most Atlassian context shipped roughly 64% more per developer. It is using that finding to argue that software agents perform better when tied to project and organisational data rather than working only from local code prompts.

Taroon Mandhana, Chief Technology Officer, AI & Teamwork, Atlassian, described organisational context as the main constraint on wider AI use in software engineering.

"The biggest bottleneck in AI software engineering isn't model intelligence, it's organizational context," said Mandhana. "Enterprises need more than isolated sessions and one-off prompts. Jira has long been the system of record for how teams work. By extending that foundation to orchestrate agents alongside engineers, we're giving teams a safe, measurable way to scale agentic workflows across the SDLC."

The release also shows how Atlassian is weaving AI functions more deeply into Jira, one of the most widely used systems for software planning and issue tracking. By placing agent management inside backlog, code review, and delivery workflows, the company is trying to make AI oversight part of standard engineering management rather than a separate layer.

Several of the products are being introduced in phases. Code Context is rolling out in open beta to paid customers. Agent loops, Standards, and AI review are in private early access. Agent Context Controls and Agent Usage Dashboard are due to become generally available to paid Jira customers in the coming months, while DX for Agentic Development is set for general availability to DX customers this quarter.

Atlassian said more than 350,000 customers use its software, including more than 85% of the Fortune 500.