Overview
- Atlassian has previewed Jira Planner, a tool that pulls a project’s code, Jira and Confluence history, and team context to draft structured technical specs in Confluence that both humans and machines can read.
- The firm also made its built-in Jira Coding Agent generally available and added a drop-down in the Jira UI that lets teams assign third-party coding agents such as Claude, Cursor and GitHub Copilot to work items.
- Atlassian shipped DX-based AI cost-management features that report token and spend data by team and estimate cost per pull request, and it plans to add local session writeback to capture developer-side AI activity.
- Analysts and practitioners say spec-driven planning aims to reduce friction later in the lifecycle by making requirements explicit, but they warn that token use and costs often move from code generation into planning, testing and business tools.
- For developers and managers the change shortens the ticket→spec→PR handoff and lowers context-switching, but it increases reliance on accurate project history and raises the risk that drift or incomplete data will produce wrong or costly automated changes.