Well... Sorry to say that, but... TESTS ARE SHIT! X-( And description isn't good either.

• description is misleading: listing 1 in the divisors, but not counting them as divisors (wtf!? Choose your side, man!)
• need random tests
• need checks on boundaries!!! the number of divisors evolves in a very erratic way, so using arbitrary values of n as it is for now allow implementations counting the divisors in a wrong way to pass: I didn't see the trouble with 1 at the beginning and failed on the small tests but passed all all the other ones while I was off by one.

It seems there's a problem with the submission tests here. The last test asks for the first number with 10 factors and expects a return of 120 but if you print n it shows that it actually asks for 9 factors and and a return of 120 meaning the algorithm for solving this will return an incorrect value.