Detecting this kind of conflict is the same as detecting conflicts involving one entity as far as the knowledge based reasoning is concerned. The difference is that the agent is not comparing something that is already in its target with what it wants to store, but it is comparing some other entity in the parameter block with what it wants to store in its target.
This means that agents need to know what entities have the potential to be incompatible with their targets. This is not a big problem since critics and praisers are the only kind of agents that can get into such conflicts because of incompatible criticisms and praises of the same entity.