Cursor has introduced a significant enhancement to its AI coding assistant with the launch of Bugbot Autofix. This new feature is designed to streamline the code review workflow by automatically detecting and fixing issues identified in pull requests. By leveraging cloud agents in isolated virtual machines, Bugbot Autofix not only proposes fixes but also applies them directly, which has resulted in more than 35% of Autofix changes being merged into base pull requests.
Since leaving beta, the feature is now accessible to all Bugbot users and represents a leap forward in automated code review processes. Bugbot’s efficacy has notably improved, with the average number of identified issues nearly doubling in the past six months and the resolution rate climbing from 52% to 76%.
Looking ahead, Cursor plans to expand Bugbot Autofix’s capabilities to include custom workflow automations that extend beyond traditional code reviews. Future developments will focus on self-verifying findings, conducting deep research on complex issues, and enabling continuous scanning of entire codebases, further enhancing the productivity and reliability of software development teams.

Collections

Cursor has launched Bugbot Autofix, an AI-powered feature that automatically detects and fixes issues found in pull requests using cloud agents in isolated VMs. Over 35% of Autofix changes are merged into base PRs. Since leaving beta, it's available to all Bugbot users. The average number of identified issues has nearly doubled in six months, and the resolution rate has risen from 52% to 76%. Future plans include custom workflow automations, self-verifying findings, deep research on complex issues, and continuous codebase scanning.

Enhancing Code Reviews with Bugbot Autofix by Cursor