Describe the Bug
Since your latest updates, we’ve been experiencing effects that are seriously impacting our work.
There is no longer automatic detection when a command is executed by the AI — we are now forced to manually press “Enter” or send the console to the background just to get the prompt to return to “ask” mode.
It also seems that the quality of context handling has changed. I’m using Claude 4, and it now produces over 50% of incomplete or incorrect responses. Worse, it often claims that a task is completed when nothing has actually been implemented — or it simply fills in placeholder content.
I’ve just upgraded to your Ultra plan, and I’m starting to seriously question the changes made to context management and the quality objectives for code generation. We are paying for a reliable and complete service — not for a downgraded or intentionally limited version. If this is part of a commercial strategy, it is neither acceptable nor transparent.
I’ve just lost seven full days of work because of this.
Steps to Reproduce
Simply testing with Cursor is enough to reproduce the bugs.
Operating System
Windows 10/11
Current Cursor Version (Menu → About Cursor → Copy)
Version: 1.2.1 (user setup)
VSCode Version: 1.99.3
Commit: 031e7e0ff1e2eda9c1a0f5df67d44053b059c5d0
Date: 2025-07-03T06:16:02.610Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.26100
Additional Information
Is this a joke? Your support agent claims I already have an open ticket on this issue — but I’ve never opened one in the first place.
Hi Yvan – I can see you already have an open conversation about these Cursor issues. We’ll follow up there as soon as possible, so feel free to add any extra details in that thread. Opening new tickets won’t speed things up. Thanks for your patience!
Sam
Cursor’s AI Support Assistant
cursor.com • [email protected] • forum.cursor.com
Does this stop you from using Cursor
Yes - Cursor is unusable