Preventing AI agents from executing destructive terminal commands

2 comments

• majesty-m 4 hours ago

  Happy to answer questions or go deeper into the design if anyone is interested.

• majesty-m 4 hours ago

  [flagged]