Even if it is the case that the first agent causes we know are causes in the same genus as their effects (like a man fathering a son, or a moving object causing something else to move) this is not what we are most of all mean by a cause, nor what we are looking for when we ask for the agent cause of something. Cajetan makes this clear in an account of univocal causality (a cause that is in the same genus as the effect):
Where there is univocation, there is not a cause and effect formally and per se, but materially and per accidens, since the form of the effect does not depend on the form of the cause. For the humanity that is in Socrates formally taken, depends neither in being nor in becoming upon the form of the father of Plato, but the humanity of Socrates, because it is this [sc. humanity] therefore depends on a father. Consequently, the humanity that is the foundation of the similitude between father and son is not in the genus of cause and effect, except materially and per accidens…
Say one wanted to speak formally about the agent cause of Socrates. The only manifest agent is the mated pair of Mr. and Mrs. Socrates. But what does this explain? By positing Socrates Sr. as a cause, we do not so much explain a particular instance of humanity, we simply posit another instance of it. There is something inadequate in this explanation which fails to address what someone is actually asking for. A series of homogeneous agents, each which gives rise to a subsequent one, is analogous to a train of cars in a pile-up accident, each one of which damages the one in front of it. Say the pile-up is ten cars long. If the man in the fourth car went home to his wife and explained the damage to his car by saying “the car behind me ran into me”, there is something at least inadequate, and even somewhat deceptive about his answer. Why so? Among other reasons, he is citing something as a cause of the accident that was no more a cause of it than he was- in Cajetan’s vocabulary, the car behind him was only materially and per accidens the cause of his damage. “To give a cause of the accident” means something other than multiplying causes that are all of the same kind, even though it is not entirely false to cite such causes as causes of damage. Yet by definition all non-universal causes are of one in kind.
“Transcendent cause” or “a cause transcending the genus of the things caused” is is a redundant pleonasm (redundant pleonasm- wink). Cause primarily means something that transcends a genus, even if we can call the causes in a genus real causes (though only in a derivative way).
 Cajetan, Commentaria in Prima pars q. 4 a. 3 no. 6 “Ubi enim univocatio ibi non est causa et causatum formaliter et per se, sed materialiter et per accidens quoniam forma effectus non formaliter dependet a forma causa. Non enim humanitas quae est in Socrate, formaliter sumpta, dependet in esse vel fieri in humantate Platonis patris, sed humanitas Socratatis, quia est haec, ideo dependet a patre. Et consequenter humanitas, quae est fundamentum similitudinis inter patrem et filium, non est de genere causa et causatae, nisi materiaiter et per accidens…”