All post-exploitation C2 agents that I'm aware of are implemented as a single rDLL or PIC blob. This means that all of their core logic such as check-in's, processing tasks, sending output, etc, are all mashed into a single executable blob. If an agent is implemented as an rDLL, then
