Cursor Acquires Graphite to Own Review

This deal shows Cursor is trying to own the full path from writing code to getting it merged, not just the editor where code first appears. Graphite gives Cursor a native home for pull requests, AI review, merge queues, and stacked PRs, which are small linked code changes that are easier to review. That matters because GitHub’s advantage is not only code generation, it is control of the approval and merge step where teams decide what actually ships.1 2 3 4 5

• Graphite was already built around the exact workflow Cursor needed. Its product combines a PR inbox, stacked PR tooling, merge queue, and AI review, and Graphite said the platform was used by hundreds of thousands of engineers. Buying that product is faster than teaching an IDE company to build review habits from scratch.2 3 4
• The strategic target is GitHub’s outer loop. GitHub already owns the repo and PR surface, and prior research on CodeRabbit shows how dangerous that platform control is for independent review tools. By bringing Graphite in house, Cursor reduces its dependence on GitHub for the review experience, even if the repos still live there.5 6
• This also fits the broader shift toward agent workflows. As more code is generated by agents, the valuable checkpoint moves to reviewing diffs, validating intent, and controlling merges. Warp has made the same bet by pulling code review into its own environment, which suggests review is becoming a core product surface, not a side feature.8 9

The next step is a bundled engineering workflow where Cursor writes the code, Graphite reviews it, and the system learns from every accepted or rejected change. If Cursor executes, it becomes less like an AI autocomplete tool and more like a full engineering control plane that can challenge GitHub at the team workflow layer.1 4 8 9