IMO, checks that can be automated, like linting or formatting, should happen before code review and have been fixed before the code is submitted for review.
That way the code review is not 'polluted' by potentially plenty of automated comments, and reviews can focus on actually review the code as it is proposed to be submitted to the code repository.
I find the use of “she” throughout your site when referring to an automated code process to be fairly strange. I suspect you’re trying to add levity and personality, but it’s an odd thing to assign gender to. Perhaps consider replacing those with “it”?
Apparently someone asked the same question - from the FAQ:
> Danger keeps getting referred to as “her/she”, what gives?
> When we were in the naming process of Danger, we went through a lot of names. In the end, it got named after Gem “Danger” McShane, who was involved in both the copy and the concept. While not a programmer herself, she’s helped out with trying to understand the personality behind the project. Danger, the project, is a female, northern UK, command-line app into 80’s punk/indie. Inspired by community efforts like the Haçienda / Factory Records.
> Understandably, applying a gender to a genderless concept comes with it’s own complications. To make it worse, female genders tend to be applied towards “digital assistants” which should raise questions on outdated-but-still-persisting gender stereotypes and patriarchical power structures. There is no answer that can completely absolve the disconnect I’m afraid.
That way the code review is not 'polluted' by potentially plenty of automated comments, and reviews can focus on actually review the code as it is proposed to be submitted to the code repository.