My thing about ExploitGym in particular is that it's testing environment varies depending on the exploit. From what I've read, it tends to favor userspace sandb

My thing about ExploitGym in particular is that it's testing environment varies depending on the exploit. From what I've read, it tends to favor userspace sandboxing inside Docker containers for standard program exploits (eg Chrome), and only uses QEMU for Linux kernel stuff (still in Docker, from what I understand). So every time I hear that an agent "broke out" of a environment, I have to question the kind of isolation used. Because escaping a sandbox or container engine is a lot more trivial than something escaping something like ESXi or Proxmox.