Is it true that all agreed upon human knowledge is based on assumptions of self-evidence?

It is based on a chain of knowledge that ultimately ends up in something that is either an axiom or is self-evident.

Not “assumption”, no. You can’t say you “know” something if it is based on an assumption.

However, large parts of this “chain” of knowledge that rest upon other knowledge are gathered through induction, meaning it is only known through probability, which is why it’s possible for such a chain of knowledge to collapse due to a weak link, in theory.

This is an insurmountable problem of epistemology.

However, a unique characteristic of this “chain” is that the longer it is, the more likely it is to be correct. That is, the more knowledge is built upon other knowledge, the more evidently correct the previous links in the chain become because it becomes less and less likely that you could build a more and more exact model of the world that actually works and still be fundamentally wrong.