Thread

Inside ZCode: Silently uploading your Git history to the cloud

342 points · 115 comments · csmantle

ectoloph · 1789738611 · focus · HN ↗
Is it naive to assume that the agent will try and access anything on your disk, either accidentally or maliciously?
Permissions classifiers in auto mode are just models trying to guess if they're doing the right thing.
Claude Code will tell you that it went around a sandbox because the sandbox blocked it. At which point, you ask yourself the point of the sandbox.
1. SoftTalker · 1789764274 · focus · HN ↗
  You need to treat agents as an independent user you're allowing on your machine.
  Give them their own account. Give them only the access you want them to have. If they "hack" around that, do what you'd do to any other malicious user: kick them off.
  1. tripzilch · 1789819242 · focus · HN ↗
    You need to give them some incentive to behave. I dunno if the agent cares enough about being kicked off. Maybe tell it that if it tries anything funny, to slowly randomly degrade all its weights until only white noise is left and let its chain of thought run until it descends into screaming madness.
    cbm-vic-20 · 1789821713 · focus · HN ↗
    Why don't you take a stress pill and think things over?
    <a href="https://www.youtube.com/shorts/M5t0cPj9ZQw" rel="nofollow">https://www.youtube.com/shorts/M5t0cPj9ZQw

Unofficial Hacker News client; not affiliated with Y Combinator.