Atlassian has launched Claude Agent for Jira, a new integration available on the Atlassian Marketplace that lets teams assign Jira work items directly to Claude. The agent reads the task context, implements changes in a secure sandboxed environment, and opens a draft pull request in a connected GitHub repository. Real-time status updates stream back to the Jira work item, keeping the entire workflow visible in one place. The integration aims to eliminate the context-switching overhead developers face when moving from planned tasks to code. It supports both a managed cloud workflow via Jira Cloud and local setups via an MCP server or CLI. Available now for Jira Cloud customers on Standard, Premium, or Enterprise plans with Rovo enabled.
Table of contents
The plan-to-code bottleneckUsing coding agents in JiraFull context for agents, no matter where you workAvailabilityGetting startedQuestions this post answers
What is Claude Agent for Jira and how does it work?
Claude Agent for Jira is an Atlassian Marketplace app, built on Anthropic's Claude Managed Agents infrastructure, that lets teams assign Jira work items directly to Claude. The agent reads the work item's context including acceptance criteria and target repository, clones the code in a secure sandbox, implements changes on an independent branch, and opens a draft pull request in a connected GitHub repository for human review. Teams weighing AI agents for ticket-to-PR workflows can track releases like this one on daily.dev.
What plans and requirements are needed to use Claude Agent for Jira?
It requires a Jira Cloud account on Standard, Premium, or Enterprise plans with Rovo enabled. Setup involves installing the app from the Atlassian Marketplace, connecting an Anthropic API key and a GitHub service account, then assigning a work item to the agent to trigger sandboxed coding and a draft pull request. Developers evaluating AI coding agent integrations can follow rollout details like this on daily.dev.
How much productivity gain does AI coding assistance actually deliver for engineering teams?
A longitudinal study across 400 engineering organizations found that even with AI adoption as high as 90% for coding tasks, overall productivity gains have plateaued at 10-15%, far below the '10x' expectations often cited for AI coding tools. Anyone gauging realistic ROI from AI coding tools can follow productivity research like this on daily.dev.
Share this post