What’s the difference between Bugbot and Agent review feature in the Cursor IDE (see image)? Are they the same thing, or do you need to run both to find any errors in the code?
Hey, these are two different features. They complement each other, they don’t duplicate each other. They run at different stages.
Agent Review runs inside the IDE on your local changes, before you push. You can run it via /agent-review in the agent panel, via the Source Control tab which compares all local changes against main, or automatically after a task if you enable it in Cursor Settings > Agents. It’s useful to catch issues early, even before you commit.
Bugbot runs at the pull request stage. It integrates with GitHub or GitLab, automatically reviews the PR on push, and leaves inline comments with suggested fixes. With autofix, it can even commit the change for you. You set it up via the dashboard.
So Agent Review is for early local checks, and Bugbot is a team gate at the PR level. You don’t strictly need both, pick what fits your workflow. Their results can differ due to different prompts, models, or context, so one can catch things the other misses.
More details: Agent Review | Cursor Docs and Bugbot | Cursor Docs
Thanks for your reply, Dean—as kind as ever.