Cursor 2.1: Improved Plan Mode, AI Code Review in Editor, and Instant Grep (cursor.com)

Cursor 2.1 adds several productivity-focused upgrades: Plan Mode now asks clarifying questions and surfaces an interactive UI so you can answer them inline, plus you can ⌘+F to search within generated plans. The editor gets built-in AI code review that inspects your changes and reports issues in a side panel, complementing Bugbot (which runs on repositories via GitHub, GitHub Enterprise Server, GitLab, etc.). Search performance is also improved—all grep commands run by the agent are now instant, and instant grep works for manual sidebar searches too, including regexes and word-boundary matching. The rollout to 2.1 users will be gradual over the next week. These changes matter because they shorten the developer feedback loop and reduce context switching: clarifying questions improve plan precision before coding, in-editor reviews surface bugs earlier without leaving the editor, and instant grep makes code navigation and pattern searches feel realtime. Support for regex/word-boundary matching and coverage across all models means the speed and search capabilities are model‑agnostic, while integration with source-control Bugbot preserves team workflows and enterprise compatibility. Expect faster iteration, fewer trivial PRs, and more continuous, low-latency assistance—though teams should still validate automated findings as part of their review process.