I got the same trouble: using static dictionary to store registers. So, clear your variables and structures before executing code! ))))

I did it, I'm not proud of my decision at all, this kata devastated me!:))
But it worked and I was too lazy to improve anything after I saw the cherished submit))
It was definitely a fun experience, but I wouldn't do it again. Good series of katas

Excellent kata! No description, just a few test cases... Have a big pleasure while solving it. Would recommend to my friends, thanks for fun!

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.

I have a problem here: gap(12, 30091, 30103) - with local testing got Null, but on attempt at site have an error, that solution find whis pair as answer. Could anyone help?
Random tests on submit completed well

Anat2017, as pointed in description, source list doesn't containt negative integers, so there is no any sence in additional filtering.