Is there any documentation for YOLO mode? I have yet to get it to fire and it’s because I’m just shooting in the dark looking at other people’s postings but I feel like I’ve missed some sort of documentation talking how you can utilize it.
I’ve got a python script And I cannot get it to fire To then test whether things are working… as I have some rules in .cursorrules to try to get it to troubleshoot if there are errors.
If anyone has a link or a video I’d greatly appreciate it:
Unfortunately there isn’t really much documentation here as everything you need to know is at least briefly explained within the editor.
When adding commands to the allow list, I believe the AI shouldn’t be able to run any other commands besides those added to that list. If the allow list is empty, the AI will run all commands on its own, unless you add some entries to the deny list, which would not be run in this case.
Let me know if you have any other questions about how this works.
So based on what I wrote about trying to get it to run:
python analyze.py
is there anything I can do to make it happen? Should it be running? What would be preventing it from running? Just trying to get it to fire that.
Thanks!
I am more or less in the same boat.
I enabled yolo yet I have no idea how to use it.
I tried in the chat stating “run LS and give me the list of files on this directory” yet the command was not ran.
I mean… HOW do you run commands sent by the ai chat?
@logichammer For certain things like python commands, the AI may not want to run these inline, as they are main programs that should usually be run from the main terminal.
The idea behind the agent running commands would be to do ‘pip install’ or ‘wget’ commands, which may be needed to setup a project once, but not to run your full applications. You can always run the code yourself, select the output in the terminal,l and do CMD+L to add that output to the AI’s context, if that is what you are trying to accomplish.
@a.j YOLO mode only works in the Composer, with agent mode turned on, as the agent composer has the ability to suggest commands to be run in a terminal and, with your acceptance, run them for you. YOLO mode just removes the need to accept the command, and allows the AI to run whatever commands it wants without confirmation.
