I'm not sure it's feasible to do on Codewars. The methods, derived from Dixon method, would only be effective for large numbers (let's say, above 64 digits), for smaller numbers other methods (such as https://en.wikipedia.org/wiki/Shanks%27s_square_forms_factorization) are faster.

Factorizing something like 2^64 would not fit into Codewars 12s test suite, whereas if I take smaller numbers, users may exploit faster methods to achieve the same task.

Don't worry, I am not working as a programmer, and probaby I will not again :D

approved at some point

Your implementation of the random tests makes it possible for the function to be passed no arguments, which contradicts the description.

The function is annotated with tuple[int, int, int, int], but you expect a list[int] in the tests. Please change the tests to expect a tuple.

If they're not flagged as spoiler, then yes, everyone can see them, here and in the homepage.

ever heard of the Spoiler button? :/

Nevermind

You should use BigInt in javascript

Which solution should be used as reference solution (which is fastest)?

performance constraints locked :)

Interesting one, the math behind this kata escaped me :)