Skip to content

Comment on A deep dive on agent sandboxes

Comments

Cool article. It makes me think about an "old school Unix" approach which might work for some use-cases.

Essentially, the untested brainstorming-only idea is:

  1. Make $HOME have 0751 permissions
  2. Assume the dev project exists in $HOME/foo and has
     0715 permissions
  3. Assume $HOME/foo/src is where all source code resides
     and has 0755 permissions (recursively)
  4. Install the agent tools with a uid:gid of something
     like llm:agent
  5. Turn on the setuid/setgid bits for executable(s) in
     the agent tools or make wrapper script(s) having
     same which delegate to agent tools
This would ensure agent tooling could not read nor modify $HOME, only be able to read $HOME/foo (top-level project directory) and its files (assuming `o+r` is the default), and could only modify files in $HOME/foo/src having `o+w` permission as well. If agent directory creation in $HOME/foo/src is desired, enable `o+w` on it and directories within it.

There is probably some "post agent use" processing that would be needed as well.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.