All prime gaps are even, but you shouldn't have to test for that specifically if your code functions properly. If your code is returning <30091, 30103> when the code is testing for a gap of 11, something is wrong.

I don't understand your question. There's no test for gap(12, 30091, 30103) and the random test generators don't seem like they'd be able to generate a test like that either, at least in C# (which I'm assuming is the language you're talking about). In any case, while 30091 and 30103 are both prime numbers, they aren't successive prime numbers. 30097 is also a prime number, and that falls in between the two previously mentioned primes.

Asking for help is not a kata issue.