However, though, as to the original statement, I had some dumber naive thing in mind, that I feel like must be wrong.
If we know that $a^{-1}p\subseteq p$, then don't we know that $a^{-1} \subseteq \mathfrak{o}$? Does this then imply that $\mathfrak{o} \subseteq a$? I was preceding by basically multiplying each side, which feels wrong with set inclusion.