📘 🏛️ Decide on AI contribution policy and document it

[e-lo]

As a contributor and reviewer, I'd like to know what our policy is w.r.t. use of AI in code submissions and if and how AI was used in the development of code submitted by a pull request in order to know how to evaluate contributions.

Question for discussion: If/how should we limit code submissions from AI?

Some thoughts:

• Some projects with high throughput limit any contributions by AI - for very good reasons as it turns out.
• We don't have very high throughput
• We DO want to make sure we aren't adding slop and that the codebase is legible and managable

Straw-person policy:

AI-assisted contributions are welcome, but contributors are expected to thoroughly review, understand, and stand behind any AI-generated code or text as if they had written it themselves. Ensure all tests pass and that any new or AI-generated tests actually validate meaningful behavior — not just satisfy coverage metrics or placate a linter. Low-quality or poorly-understood contributions will be closed without extended discussion.

[lmz]

I like the straw-person policy and I'm fine with starting with that. 😄

[e-lo]

Q: Should we require disclosure as well?

[lmz]

Yes, I think it doesn't hurt.