I think it's okay to keep different types as inputs but the description is really off

Given a list of integers

in context of Haskell seems like a literal requirement for the arg to be [Integer] type

There should be Haskell-specific description clarifying that input is a list of Integral derivative

wouldn't it be fine if simply defining the type in the default code? it's quite odd for the solver to specify the type when they don't .. know what it should be

It's solvable using typeclasses

But yeah it does put more of a challenge compared to other languages