Cursor Adds Bugbot Autofix — Automatically Fixing PR Issues
Cursor's Bugbot now includes Autofix, using cloud agents to automatically fix issues detected in pull requests. Over 35% of Autofix changes are merged into base PRs.
Cursor announced on February 26 the addition of Autofix to Bugbot, its AI-powered code review tool. The feature uses cloud agents running in isolated VMs to automatically generate fixes for issues detected in pull requests.
How Autofix Works
When Bugbot identifies problems in a pull request, Autofix spins up cloud agents on dedicated machines to test changes and propose fixes directly on the PR. According to Cursor, over 35% of Bugbot Autofix changes are currently being merged into base PRs — a notable adoption rate that suggests the fixes are genuinely useful rather than noise.
Bugbot posts a comment on the original PR with a preview of the proposed changes. Developers can merge the fix using a provided @cursor command. For teams that want a more hands-off workflow, Autofix can be configured to push changes directly to the branch without requiring manual approval.
Getting Started
Autofix is available now through the Bugbot dashboard. More details are available in Cursor’s official announcement.
The Bigger Picture
This release continues Cursor’s push toward end-to-end AI automation of the development workflow. Bugbot already handled detection; now it handles remediation too. Combined with the recently launched cloud agents with computer use and CLI improvements, Cursor is steadily building out an ecosystem where AI agents don’t just flag problems — they solve them.
Related Articles
Cursor Adds Subagents, Skills, and Image Generation - Major Agent Enhancements
Cursor implements subagents, agent skills, and image generation. Parallel execution, improved context management, and dynamic skill application enable handling of more complex, long-running tasks.
Cursor Launches Long-Running Agents in Research Preview - Autonomous Execution for Complex Tasks
Cursor introduces long-running agents capable of autonomous planning and execution without human intervention. Available for Ultra, Teams, and Enterprise plans.
Cursor 2.5 Brings Plugin Marketplace, Sandbox Network Controls, and Async Subagents
Cursor releases version 2.5 with a plugin marketplace for extensibility, granular network access controls for sandboxed environments, and asynchronous subagent execution for parallel processing.
Popular Articles
Claude Code v2.1.93 Released - Deferred Permission Decisions, Flicker-Free Rendering, and More
Anthropic releases Claude Code v2.1.93 with deferred permission decisions for PreToolUse hooks, flicker-free rendering option, PermissionDenied hook, and named subagent typeahead support.
Claude Code v2.1.92 Released - forceRemoteSettingsRefresh, Bedrock Setup Wizard, and More
Anthropic releases Claude Code v2.1.92 with forceRemoteSettingsRefresh policy setting, AWS Bedrock setup wizard, /cost command improvements, and numerous bug fixes.
Claude Code v2.1.84 Release - PowerShell Tool Preview and Environment Configuration Enhancements
Claude Code v2.1.84 introduces PowerShell tool for Windows, new environment variable overrides for model selection, idle session handling improvements, and various stability fixes.
Latest Articles
Claude Code v2.1.93 Released - Deferred Permission Decisions, Flicker-Free Rendering, and More
Anthropic releases Claude Code v2.1.93 with deferred permission decisions for PreToolUse hooks, flicker-free rendering option, PermissionDenied hook, and named subagent typeahead support.
Claude Code v2.1.92 Released - forceRemoteSettingsRefresh, Bedrock Setup Wizard, and More
Anthropic releases Claude Code v2.1.92 with forceRemoteSettingsRefresh policy setting, AWS Bedrock setup wizard, /cost command improvements, and numerous bug fixes.
Claude Code v2.1.91 Released - MCP Tool Result Persistence and Improved Edit Tool
Claude Code v2.1.91 introduces MCP tool result persistence override, improved shell execution controls, and enhanced Edit tool efficiency.