add LLM use policy and contributing guide - #442
Conversation
brandonwang15
left a comment
There was a problem hiding this comment.
lgtm
Separately though, I wonder if, since this is public, should we eventually include a note in our README about our own norms/principles around writing and reviewing AI-generated/assisted code?
We probably require more internal discussion / thought to figure out what that is ourselves, and how we want to present that, but it might be a good thing to have clarity on for something public-facing.
Also PRs are public on this repo too, just FYI, so fwiw this is technically visible already
|
@brandonwang15 do you think that we want to add any AI principles to this repo, like: https://github.com/catalyst-cooperative/pudl?tab=contributing-ov-file#llm-use-policy? I think as written, this file also provides relevant design standards for contributing to this repo whether you are an LLM or a human. Would it make sense to rename this to a "CONTRIBUTING.md" file so that it is less tied to Claude specifically (but if using an LLM, you could point it to this file to understand how to contribute)? |
I'd be pro adding something like that, I like the PUDL writeup. And yeah renaming |
|
@brandonwang15 @sarahchute I made some updates based on your feedback, adding an LLM policy (copied from PUDL), and a section to the PR template to describe LLM use. I also renamed the file |
Purpose
Adds a markdown file that provides instructions to LLM models on how to assist with code changes in this repo.
The main question for reviewers is: should we include this in a public repo? There is nothing private/sensitive in the file (it is all derived from the codebase itself), but we just haven't established a policy for this yet.
Also updates the README with a LLM use policy and the PR template