013-greatest_common_divisor does not guarantee that result is the gcd

[jesyspa]

The postcondition does not relate the result to the inputs at all.

[alex28sh]

We could prove that gcd divides both a and b, like that https://gist.github.com/alex28sh/fc2608027fdb923b1f524fe7ccfad5d9

[alex28sh]

Though it requires some changes in code, it preserves the idea with modulo division

[jesyspa]

I think that would be an improvement, though proving it's the greatest such number would also be interesting. I'm not sure whether it's best to express divisibility by this kind of existence statement, or by n % k == 0, by the way.