I'm trying to prove that in an n-party setting, where each party has a private value, the dominant strategy is always to reveal it. I'm assuming that parties only care about monetary payoffs and therefore I can directly model a utility function that is known and leveraged by the mechanism I'm designing.
Is there any problem with designing a mechanism when the utilities are known? I know that in general the preferences are not strictly monetary and are not well-defined, so I couldn't find any papers that design a direct mechanism, in which truth is a dominant strategy, but where the utilities are known.
It sounds to me like my problem is actually easier because of this assumption, but still correct.
This is known as the Implementation Problem. Since you are asking for dominant strategies, the Gibbard-Satterthwaite theorem applies. With at least three people, and using unrestricted utilities, and there exist preferences so that your choice rule will pick out any given actual outcome, then the choice function is dictatorial. That is, there is someone who gets his favorite outcome no matter what others think. If preferences can be restricted, something called the Groves-Ledyard mechanism can work, and if you allow weaker forms of agreement, such as Nash or Subgame perfect equilibrium, there are more choices. See chapter 10 of Osborne and Rubinstein's A Course in Game Theory.