Connect Bitbucket Cloud to your AI tools with Atlassian MCP
Atlassian has expanded its MCP server to connect Bitbucket Cloud to AI tools via OAuth, enabling assistants to access code, pull requests, and pipeline data directly within development workflows.
Developers rely on Bitbucket Cloud for repositories, pull requests, and CI/CD pipelines to validate changes before deployment. Without a direct link, AI assistants previously required manual input of this context. The updated Atlassian MCP Server now supports OAuth authentication in addition to API tokens, simplifying the connection process for compatible AI clients. Once established, the connection grants assistants access to code, commits, pull request discussions, and pipeline results, reducing the need to switch between tools.
The integration allows AI assistants to perform tasks beyond retrieval, such as managing pull request comments, updating reviewers, locating commits, and identifying related pull requests. This enables assistants to assist with implementation and code review rather than just providing information. For example, an assistant can use linked Jira issues and Confluence pages to guide code changes or analyze failed pipelines to suggest fixes, streamlining troubleshooting and implementation.
Atlassian MCP makes Bitbucket Cloud data accessible in various AI clients, including IDEs like Cursor or VS Code and assistants such as Claude or Codex. The connection respects existing Atlassian and Bitbucket permissions, ensuring secure access. Users of Atlassian MCP v1 must upgrade to v2, which may require reinstalling plugins or connectors and signing in via their Atlassian account.
The expansion aims to support AI-driven development workflows by exposing more Bitbucket capabilities through APIs and MCP, including pull request blocking status, merge queues, and deployment history. Atlassian invites feedback on preferred AI-assisted workflows via the Bitbucket Cloud Community, emphasizing ongoing improvements to integrate AI into development processes.