That Ruby solution was invalidated some time ago.

Fixed.

I just solved the Haskell version and everything was cool. I'll mark this as resolved.

Fixed Haskell

If there are other language versions with the same issue, please raise a separate issue with a list of all of them.

In general float comparisions can be problematic like that because the order that you do operations will change the results so it can throw the rounding either way in close situations. It is better off, but more complicated to directly do a float comparison.

Be careful about your implementation of foldr'!

Fixed.

Resolving issue for approval

push n - Pushes an integer onto he stack.
It is in the description all along.

lastDigit is monomorphic and has type [Integer] -> Integer. So I see no option, how

property (\(NonNegative x) (NonNegative y) -> lastDigit [x, y] == x ^ y `mod` 10)

could possibly instantiate x and y with Int.

Did you find the problem? I am afraid that the Unit tests crashes the run and has no time to say something... which is bad but the kata has no power on that. It would be the server that should warn with something like "Crash"...
You pass the "Example tests" where there are no problems but not all of the fixed tests when you "attempt". Usually I give as advice to print the input but in this case you would not even see the log.

I tried your solution and find kind of a flaw in your code. Think of a possible corner case. Please tell me when you succeed. Cheers!!