Can we have an option to turn off accept/reject after cursor made changes to the editor? It feels redundant since we already track changes from git.
1 Like
I would 100% agree with this. IMO it just gets in the way of my workflow. I already review all my and Cursor’s changes in the git diff viewer and from there I can undo any change.
I would love to see a toggle in user settings for “Auto Accept”.
Furthermore, the Accept/Reject buttons are bugged when opened in the diff viewer. They do not do anything.
1 Like
@shawon.kamal thank you for the feature request. Cursor team will consider the feature request.
From the current behavior you may ignore approve/reject and with new requests the changes of previous request will be accepted. The UI offers users the ability to reject or change things in cases AI made mistakes.
You can also accept all at end for now.
@belmin Could you please file a separate detailed bug report where your Diff viewer Accept/Reject buttons are buggy. It’s not a common issue and may need to be checked in detail.
1 Like
One of the worst parts of the Cursor user experience. Please make this an optional feature. As others have said, review is done in Git.
2 Likes
Any update on the feature request?
hi @shawon.kamal currently you do not need to accept changes, as it auto accepts on commit and on any follow up request. please let me know if this is working for you.
If there are more users requesting this feature we may increase the priority for your feature request.
I very rarely hit accept, just keep making changes until I commit to Git, then all the changes are accepted in Cursor.
1 Like
Yes I also never click Accept as it works with a commit per settings.
1 Like
I find that Cursor frequently has trouble with larger edits, and if I don’t click accept regularly, it won’t move on to the next section of the problem. As a result, I have to babysit the accept button.
Literally created an account to complain about this terrible feature. About to move back to VS Code. It makes the workflow completely unusable.