doc: Add docs for reviewing xdp code

[swick]

The intent is that those files should be useful for humans and LLM agents at the same time.

This is an experiment, and keeping those files up-to-date will require some work and the value of them is not entirely clear. Still worth a try in my opinion.

[whot]

a few minor nitpicks but i think this is a pretty good start, thanks

[swick]

Thanks, going to merge in a few days if there are no objections.

[tytan652]

Feel free to mark it as off-topic, but I would have preferred a ban on AI/LLM tools (alike to some GNOME projects).

[tytan652]

I think before merging this, the AI policy of the project needs to be discussed (and not implicitly decided by one committer with an AGENTS.md).

#1963 (comment):

If anyone want to add a LLM policy, it needs consensus from committers.

I think this PR falls under the same consensus requirement.

[swick]

I don't see why documentation on what we consider good practices for reviewing code requires a policy on LLM generated content, and neither do I see why an AGENTS.md file does. All of those things can be useful no matter if you use LLMs to generate content or not.

[tytan652]

All the actual contributor documentation is in RST alongside of the rest of doc (which turns into the documentation website, even the CONTRIBUTING.md redirects to it).

With that in mind, it can make the choice of markdown look like it is mainly related to be used alongside the AGENTS.md.