Atlassian Corporation (NASDAQ: TEAM) announced new capabilities in Jira designed to integrate AI agents into software development workflows, including tools for planning, task assignment, and monitoring across the software development lifecycle.

The announcement includes Jira Planner, a feature that draws on Atlassian's Teamwork Graph — a data layer connecting work, teams, goals, code, and knowledge — to generate technical specifications in Confluence. According to the company, agents using Teamwork Graph context produced 44% more accurate results while using 48% fewer tokens in internal benchmarking.

Paid Jira Cloud customers can now assign work items directly to third-party coding agents including Claude Code, Cursor, and GitHub Copilot at no additional cost. A built-in Jira Coding Agent, also included in paid plans, converts work items into pull requests without requiring a local development environment.

Additional features announced include Jira for Slack, which converts Slack conversations into structured Jira work items; Loom video prompts that translate screen recordings into agent-ready instructions; and an agent visibility dashboard showing the status of active coding agent sessions.

Atlassian also introduced autonomous workflow automation within Jira's existing rule builder, an agentic engineering project template for faster onboarding, and a DX AI cost management report that maps AI spending across tools such as Claude, Cursor, and GitHub Copilot to engineering outputs including an estimated cost per pull request.

Most features are available now for paid Jira Cloud customers. Jira Planner is in early access preview, Microsoft Teams integration is in early access, and Codex support is listed as coming soon. The DX AI cost management report is available for Atlassian DX customers.

According to an Atlassian DX longitudinal study cited in the press release, AI usage among engineers has increased by 65%, while developer velocity gains have remained at approximately 10%.