Can u please provide any reference to connect golden reference model (verilog) to scoreboard for comparison?

In reply to dave_59:

Considering the first suggestion, to duplicate the agents. If the data inputs are of rand type, how do we ensure that randomization shall generate same values from both agents ?