Atlassian Feature Trails

This post introduces Rovo Dev Code Reviewer, an AI-driven agent integrated into Bitbucket to automate parts of the code review process. It uses a zero-shot structured prompting approach with LLMs (Claude 3.5 Sonnet) and a multi-stage quality check (LLM-as-a-Judge for factual correctness, ModernBERT for actionability) to provide feedback on pull requests. The system augments human reviewers, aiming to reduce PR cycle time and human-written comments. Internal evaluations show a 30.8% reduction in PR cycle time and a 35.6% decrease in human-written comments, with a 38.70% code resolution rate for AI-generated comments.

This post introduces shadow IT controls for Bitbucket Cloud, expanding the 'product requests' feature to allow workspace admins to prevent users from creating new Bitbucket workspaces without approval. This builds upon existing controls for Jira, Confluence, and JSM.

This post details how Atlassian Marketplace Partners (JXL, ScriptRunner, and codefortynine) implemented data residency for their apps. It covers their motivations, the regions they support, and their approaches to realm pinning and realm migration. Key technical aspects include infrastructure setup in different AWS regions, managing region-specific base URLs, handling Connect lifecycle requests for region identification, and migrating customer data and configurations between regions. Challenges highlighted include rearchitecting infrastructure for multi-region support, ensuring centralized logging and visibility, scaling testing infrastructure, and handling edge cases in realm migration, particularly when source and target regions are the same (e.g., EU and Germany).

The 'Annotate' button for git blame functionality on Bitbucket Cloud UI has been renamed to 'Annotate/Blame' to improve clarity and discoverability for users. This change has resulted in a 15% increase in feature usage.