Why is this a good assumption?

The fact of the matter is, "morality" as we've historically tended to think of it, is a logical dead end.

"Goodness" and "Badness" are based purely in the desires and feelings of individual sentient beings. And as such, there can be no universal morality, as long as there exist more than one sentient perspective in the universe, and as long as those different sentient perspectives have the potential to have conflicting interests.

The reason why moral actions often "feel" good is because, as humans, doing the "moral" thing usually means doing the thing that will gain acceptance from our community. And since we are social creatures, dependent on our community, we are hard-wired to feel good when we believed we've gained the acceptance of our community.

And the reason why the trolley problem gives us mixed feelings, is because we are aware that, no matter which option we choose, someone in our community might reject us for our choice.

We are not afraid of the potential "immorality" of the choice per say, but are rather just afraid of the potential social consequences.

In the absence of a true, "natural" morality, the best thing we have to stand in for it is community approval, or as Remarkl mentioned, our legal system. Which is, in fact, just a written approximation of community approval.

But of course, humans are just one brand of sentient creatures, and we a pretty diverse bunch. So any rules and expectations found among human social groups will mostly just constitute a tyranny of the local majority, rather than revealing any universal "moral" truths, or even any universally human truths.

The one exception, perhaps, is the basic rules of facilitating cooperation, like appealing to evidence, keeping promises, and punishing antagonistic behavior. Establishing certain norms of transactional cooperation would probably be useful regardless of culture or species.