There are four possible stereoisomers, two enantiomeric pairs, each having either cis- or trans- configuration. If you use the cyanate route you will always end up with the trans form. If it's d-, l- or dl-trans-4-MAR you get depends on the optical purity of your starting material only, there is no way you could racemize or isomerize anything on your way to the product, as the stereochemistry is set as soon as you choose your starting material.
Some clarifying pictures: https://www.thevespiary.org/rhodium/Rhodium/chemistry/4-mar.stereoisomers.html (https://www.thevespiary.org/rhodium/Rhodium/chemistry/4-mar.stereoisomers.html)