Rovo Dev

Overview

Freshness note: AI products change rapidly. This profile is a point-in-time snapshot last verified on July 6, 2026.

Rovo Dev is Atlassian’s developer-focused AI agent, separate from the broader Atlassian Rovo workspace layer. Core Rovo is about search, chat, Studio, and agents across organizational knowledge. Rovo Dev is specifically for software delivery: planning code work, generating changes, reviewing pull requests, turning Jira work items into implementation sessions, and automating repetitive engineering tasks.

The obvious audience is teams already deep in Jira, Bitbucket, GitHub, and Confluence. If your requirements, acceptance criteria, deployment context, and project history live in Atlassian systems, Rovo Dev has a stronger native context path than a generic coding assistant bolted onto the side.

Key Features

Rovo Dev spans the terminal, IDE, Jira, Bitbucket Cloud, GitHub, and MCP. The CLI can analyze codebases, generate code, refactor, debug, create tests, write documentation, action Jira work items, and create Confluence content without leaving the command line. Atlassian also documents memory, saved prompts, tool controls, subagents, server mode, worktree mode, Agent Skills, and model switching.

The IDE path runs through the Atlassian extension for VS Code and VS Code-compatible IDEs such as Cursor. It supports workspace-tied sessions, slash commands, credit-usage visibility, and repo-aware prompts. In Jira, Rovo Dev can generate code from a work item, select one or more connected Bitbucket Cloud or GitHub Cloud repositories, create separate sessions per repository, and open a draft pull request when requested.

Code review is a major part of the product. Atlassian positions Rovo Dev for Bitbucket and GitHub pull-request review, including checks against common errors and Jira acceptance criteria. Bitbucket Pipelines support is currently beta and can summarize deployments or analyze failed build steps.

MCP support is worth treating carefully. Atlassian says Rovo Dev CLI and IDE process data on the user’s local device, automatically connect to the Atlassian MCP server, and can connect to third-party MCP sources with explicit user approval before data is transferred outside the Atlassian site.

Strengths

Rovo Dev’s strongest advantage is workflow fit for Atlassian-heavy teams. A Jira item can carry the plan, acceptance criteria, connected repositories, and review context into the implementation path instead of relying on a developer to manually copy the right context into a prompt.

It is also useful for teams that want one AI layer across planning, implementation, review, and delivery telemetry. The CLI and IDE surfaces cover hands-on coding, while Jira, Bitbucket, GitHub, and Pipelines give it places to operate around the code.

Limitations

Rovo Dev is not a general free coding toy. Rovo Dev Standard is a paid product, has separate packaging from Rovo, and uses Rovo Dev credits. There is a 30-day trial, but no generally available free tier. Atlassian also documents selected limited-credit rollout for some eligible Jira sites, so teams should verify what their site has actually received before planning usage.

Teams need to understand credit usage, site activation, feature toggles, and where beta surfaces such as Bitbucket Pipelines fit into policy. The CLI and IDE are also not available during a Rovo Dev Standard trial according to current support docs; paid Standard access is required for those surfaces.

The other limitation is ecosystem bias. That bias is the point if your team runs on Atlassian, but it is friction if your planning and review process lives elsewhere. Rovo Dev is most compelling when Jira, Bitbucket, GitHub, and Confluence are already central to delivery.

Practical Tips

Start with review and Jira-to-draft-PR flows before letting Rovo Dev own larger implementation work. Those workflows have natural review boundaries and make it easier to evaluate whether the tool understands your acceptance criteria.

Keep repository instructions and Jira work items precise. Rovo Dev can pull context from Atlassian systems, but it still needs clear scope, constraints, and test expectations. For MCP, decide which third-party sources are approved before developers start wiring their own context paths into critical repos.

Watch credit usage early. The $20/developer/month Standard plan includes a monthly credit allowance, and extra usage is billed per credit. Agentic coding can burn through credits quickly when tasks are broad, multi-repo, or review-heavy.

Verdict

Rovo Dev is a serious addition to the coding-agent landscape because it connects developer AI directly to Atlassian delivery systems. It is best for software teams that already use Jira as the source of work truth and want AI assistance across planning, implementation, review, and delivery follow-through without losing explicit human control over the final merge.