Whyyyyy does the terminal in the chat stall so much? It’s so random—sometimes it works, but the next three times it doesn’t. It’s always skipping and continuing.
And why doesn’t it auto-run YOLO commands even after whitelisting them?
Why is using Gemini 2.5 so good and so bad at the same time? It keeps spitting code into the chat when it should be populating the files. Ugh.
I think Gemini needs a little buffing from the Cursor team because it’s becoming my main model. Its really close.
Have you found a way to mitigate terminal command stalls? I have written a couple of monitoring scripts but they seem intermittent and generally ineffective.
I suspect a memory leak in cursor.ai.
Running Windows 10 with 256GB RAM, I was modifying code when cursor.ai suddenly threw errors.
The system became extremely laggy, completely froze, and Task Manager stopped responding – all while memory usage skyrocketed to over 200GB. Even during model training sessions, I’ve never seen such high memory consumption.
I had to reboot to resolve it. After restarting, cursor.ai returned to normal (500MB memory usage). Recommendation: If you experience similar freezes, check your memory usage immediately.
