Phoenix `phx.new` generator and AGENTS.md

@chrismccord : I just saw the Extract AGENTS.md from Phoenix.new into phx.new generator commit to the phoenix project.

My initial shotgun all-over the place unorganized reactions:

  1. I like the idea of the community having a set of AI instructions that works well for elixir projects, which brings together experimentation from all parts of the community… As someone that just wants to “be productive” and doesn’t want to waste time “prompt engineering” an AGENTS.md file from my own frustrations… seeing what others have done/shared is nice.
  2. The merged AGENTS.md reads like an “Elixir + Phoenix + Ecto” best practices or a conventions document… Phoenix has great guides, Credo is like a best practice enforcer.
  3. Is the generator a place for this? Or would a separate community contributed “Phoenix AI Suggestions Package” (maybe credo?) that usage_rules uses to add to AGENTS.md be a better place?
  4. Is there a Phoenix project’s AI Policy, like Ash Framework: AI Policy - #6 by zachdaniel ?

I’m just curious about what the community thinks about elixir plus AI agents instructions in general.