Some of the stuff in PEP8 is pretty subjective. What's the state of play for fixing that with an automated tool? Does is spit out advisories, or has some effort been put into the Small Matter Of Programming of interpreting the intent of code? I presume nothing can be done about

    x += 1    # increment x

You can pass flags to pep8 to ignore specific errors. For our project, we have:

--max-line-lenght=120 --ignore=E128

Came to say the same. A call to flake8 is a very useful step in most of my Git pre-committ hooks.