[staticassertion]

I strongly suspect the same thing - that they weren't using agents at all in the reports we've seen, let alone agents with instructions on how to verify a viable attack, a threat model, etc.