I know that enabling Shadow Workspace with Yolo mode is the way to encourage as much autonomous behavior, with accuracy, on Cursor.
How does the new agent mode in Github Copilot measure against this?
Hm i couldnt find an option for “yolomode” in the Copilot agent. But that it has unlimited premium request is a standalone feature.
Still not as good as Cursor.
Isn’t there still some rate limiting for premium models?
100% not been playing whole day with agent. No limits so far
I am finding that it’s plan of execution is slower but better, it “thinks before act”, whereas Cursor yolo just starts spitting out the code most of the time duplicating the structure. Still early days, but looks like Cursor needs to reinvent yolo.