This is more of a general question, so apologies if it’s in the wrong place.
I’ve been a big fan of Cursor for almost a year now; but in the last ~3 months the quality and usefulness that’s decreased such a phenomenal extend… almost no AI features work now (interpreter, Copilot++, most of the chat context/docs crawling not working properly/consistently/debuggable… it’s really starting to become quite the slog.
Multiple requests for help in the forum just go unanswered - have the team moved to email support now? Inactive?
Would be great to have some ideas about what’s going on/how all of the issues the people on the forum are experiencing are progressing, so we can understand whether we should keep using Cursor.
Chris has been a really version of VS Code - an IDE that I am sure many of us really enjoy - with AI features that (used to) work! Really hoping we get some clarity soon.
Thank you for the feedback! We’re a small team that’s a bit swamped with support requests (while trying to improve the product in new ways), so please bear with us - will try to be a bit more responsive on the forum.
Would love to hear about the specific issues you are facing, so I can dig further into what’s going wrong! What have you been seeing with Copilot++ and the chat context/docs?
For me one of the most annoying issue is the broken markdown formatting on chat. Sometimes it works great, sometimes it does not render the codeblocks and sometimes it just prints a single letter per line.
As a user, I believe it’s important to recognize that building features into Cursor presents unique challenges compared to conventional software development. LLM technology is fundamentally probabilistic; every model has its own quirks, making it difficult to find a common prompt that works for every situation and edge case. I advocate for more understandings.
