I've started building AI agents for offensive security. One specific issue that I found interesting to dive in a little bit is sandboxing and access control. Which might help protect from RCE and unauthorized access -->
gr1m0ire.xyz/articles/san...