I’ve been using the new feature that lets you add commands to the .cursor folder, and just by using the name of the new command, you can save a lot of time. But it’s not only a time saver. I think it understands what you want very well, and when you combine those commands, you get a powerful way to create new plans, solve problems, understand architecture, or even explain ideas.
I also include some warnings when creating the instructions for the commands, and those warnings are taken into account by the model.
Thanks a lot, Cursor Team!