I don't think he was serious with that solution

My hero

Oh boy, needles to say, hardcoding is not a great idea.

Would you have enough 'dots' (RAM, patienсe etc) if cat can jump over 100 (or may be 1e+100) characters ?

Oh! I like this!!

This is the most thorough solution to this problem.

That string will be replaced at index x which will be the input number with the second string.

There's no mention that negative index should be handled. In fact the fact that you can use negative index in Python is because it's a sugar: in most other languages the valid indices are always from 0 to len(string) - 1.

Needs Random Tests.

Approved

Kotlin translation is ready!
https://www.codewars.com/kumite/5bac48e32ddd484f3100014f?sel=5bac48e32ddd484f3100014f

I don't think I understand what you're getting at.
The random tests are present and they generate Assert.AreEqual() statements that are equivalent to the JS ones.
Perhaps you are referring to each of the random test cases not being listed individually in the test runner output?
What do you even mean by 'real' C# code? I'm new to codewars and more than happy to learn if the criticism is constructive, but this isn't a constructive or sufficiently informative comment.

What happened to your random tests?
You cannot translate languages "one-to-one" :/

Could you please change your tests for the better please? :)
You know, more like 'real' C# code? :)