I am hesitant to allowing AI Agents full access to my laptop and want to isolate them in some sort of container or sandbox. I have used them a little and the allow/deny when they want to run scripts is worrying because the scripts take me a few minutes to review to determine if they are safe each time. This is one of the reasons I want to look into this isolation.
If you have done already sandboxed your Agents I would like to hear what steps you took to implement this.
How have you isolated AI Agents for development?