Sure! For example, you could prove in zero-knowledge that you know the prime factors of $n$, where $n$ is an online RSA challenge (i.e. something we know for sure you did not create yourself).
Such a zero-knowledge proof would look as follow: you commit to the two prime factors $p, q$. You prove that the commitments contain primes, and you prove that their product is $n$.
This is actually a relatively common zero-knowledge proof in cryptography, but in a different context: it is used to prove, when you generate an RSA key $n$, that you generated it correctly and that it has the right structure. The seminal paper describing a zero-knowledge proof for this task is this one.
In general, for any cryptosystem which you have broken, you can always be challenged to e.g. prove that you know the plaintext of a challenge ciphertext, or prove thar you know the secret key associated to a public key. For essentially all cryptosystems in use, this statement is an NP statement, hence it admits a zero-knowledge proof of knowledge (if one-way functions exist).