i d like to wait for all code finish, but i just wait. what you people do?
always running two cursor sessions, preparing the next prompt for the other session while Agent is running. Perhaps that’s why I hit the rate limit so fast
1 Like
Plan, look for issues, read the process of the ai to spot issues, Look for issues specifically in changes the ai have made since I committed last.
1 Like
I have a book in my desk that I read, I’ve made more progress on this book than I ever had by reading between prompts
1 Like
I use up to 21 instances of cursor at a time, so just checking outputs and giving them more work takes all my free time.