No problem. We can have AI do it.

And the side benefit is that we could summarily execute one every once in a while for failing to write secure code.