There are rounding problems with your test cases.

I think it should state explicitly what it expects from a non single digit input.
Otherwise, you need to submit at least once to know.

This was an issue with SQL katas that's been fixed a couple of hours ago. Try again.

I solved this kata, but it took me too much time because I took the same approach as you, just solving the equations and trying to reduce the search space, which is too slow... the non spoiler hint is that you need to consider the hint given in the Kata...

The meaning is: do not test each number for primality individualy, use a method that allows one on computing groups of primes (e.g. Sieve of Eratosthenes)

Just did this kata and I stil got the wrong code.

What's the reasoning behind this test? It clearly differs from the definition of "rounding" othe standard library, and there's no clear definition on the kata description that entails it.

Remember that these are spaces, and they are funnily rendered in HTML... check if both tests use exactly equal strings ;)

Something is wrong with the description. It says we need to compute S = S(0) + S(1) + ... + S(n), but in fact, it wants S*4.

I don't know about other languages but doing it in Java was really frustrating.

1. The tests test for exceptions thrown but then don't compile if you throw them.
2. It doesn't specify how to use the OutputStream, I needed to test a couple of methods until I found one that matched the way tests work
3. It's not enough to detect duplicate labels before running, because some tests check both if you detected the duplicate label AND the expected output AFTER running the INVALID program!!!
4. It refers to a readLine method that doesn't exist in the InputStream class

Well, people may find out automatic boxing/unboxing ;) that may be useful.

Test cases for python are broken. I get "Unknown error" half of the time when running it.