Sure, but the tokenization should cause problems for the encoding and cipher tasks as well? In addition to that: The odd/even starting letter task should be sol

Sure, but the tokenization should cause problems for the encoding and cipher tasks as well? In addition to that: The odd/even starting letter task should be solvable by only considering tokens since the first letter should fall on the beginning of the token.