[Feature]: consider having doc-and-style use "commit-directly: true" by default

[iantaylor-NOAA]

Is your feature request related to a problem? Please describe.

@e-perl-NOAA added commit-directly: true to the r4ss workflow for doc-and-style following a suggestion from @k-doering-NOAA and I love it: https://github.com/r4ss/r4ss/blob/spr_split_tables/.github/workflows/call-doc-and-style-r.yml#L21.
The changes was made after I messed up the workflow by naively having it apply to pull requests (which would have required an extra pull-request to the pull request). See also #135 and discussion here: NOAA-FIMS/FIMS#660.

Since 2021 I've now merged exactly 100 pull requests created by the doc-and-style workflow in r4ss, and never NOT been happy with the changes. Using commit-directly: true avoids all the extra work and notifications associated with merging a pull request while still providing plenty of opportunity to review the changes by looking at the commit that the GHA adds.

Describe the solution you would like.

Either add commit-directly: true to the workflow or give the user an argument to choose to add it. If there are branch protections, then committing directly to main would be a problem, but in that case everything will come through a pull request and the changes could be committed directly to the branch in the PR.

Describe alternatives you have considered

Keep the status-quo workflows and add a note in the documentation that the user could easily make the change manually.

Additional context

No response

[k-doering-NOAA]

Thanks for the feedback, @iantaylor-NOAA !

I think this option is included in the use_doc_and_style_r() function, but it is not well advertised. Here's some thoughts on how to make it more well known:

1. Specifically list the commit-directly option on the readme.
2. Add examples to the documentation showing how to set up the commit-directly option
3. We could change the defaults if others agree. I personally like the ability to look over a PR and it is a more conservative option to start out with - but perhaps others would agree this would be better!

[k-doering-NOAA]

Also, I opened #137 for the pull request problem you experienced, thank you for reporting!

[iantaylor-NOAA]

The how_to_commit argument is totally fine. I'm sorry I didn't take the time to look at the existing options.
Based on the dependency graph, I think we could just use word of mouth to spread the idea of using that switch, so feel free to close this issue.