Moltbot can be safe enough, but only if you treat it like a privileged system you are operating, not like an app you are trying. A normal chatbot mostly risks privacy. An agent risks privacy plus action.

Honest question, how are people trusting this already? I keep seeing it everywhere this week, but it is basically a tool that can touch files and run commands... The hype feels fast for something that can wreck your machine if you install the wrong thing, so just why and how

Betteridge's law strikes again!