Add AI policy to contributors guide and update PR template by nickbianco · Pull Request #4467 · opensim-org/opensim-core

nickbianco · 2026-10-05T18:37:34Z

Fixes issue #4420

Brief summary of changes

Updates CONTRIBUTING.md to include a new section "AI/LLM Usage Policy" which lays out guidelines for using AI or LLM coding assistance tools when creating pull requests. Inspired by the AI contributing guides from Homebrew and LLVM.

Also updates the pull request template to encourage contributors to disclose any AI/LLM usage.

Testing I've completed

Looking for feedback on...

CHANGELOG.md (choose one)

no need to update because...update to contributing guide.

This change is

nickbianco · 2026-10-05T18:40:54Z

@tkuchida @carmichaelong @aymanhab @adamkewley

Happy to hear any thoughts or suggestions on this!

adamkewley

@adamkewley reviewed 2 files and all commit messages, and made 1 comment.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on nickbianco).

aymanhab

@aymanhab reviewed all commit messages and made 1 comment.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on carmichaelong and tkuchida).

tkuchida

Looks good. I had a few (human-generated) suggestions, but this is mostly forward looking so I expect it will evolve regardless.

tkuchida · 2026-10-06T22:17:51Z

 ### Looking for feedback on...

+### AI/LLM usage
+<!-- Please disclose any AI or LLM tools used to create this pull request.-->


Suggest "Please disclose any AI or LLM tools that were used to create this pull request, how they were used, and what quality checks were done" (or similar) since the reviewer will want to know more than what the tools were. Also, nit: there's a space before "P" but not after "."

tkuchida · 2026-10-06T22:18:16Z


 ### Looking for feedback on...

+### AI/LLM usage


Suggest "usage" -> "use"

tkuchida · 2026-10-06T22:22:34Z

+Contributors to OpenSim are welcome to use Artificial Intelligence (AI) or Large Language Model (LLM) coding tools to create pull requests. However, all pull requests must be created with a human in the loop. The contributor is always the author and is fully accountable for their contributions. Contributors should be confident that their submission is high quality and worth the (limited) time of the maintainers.
+
+- You must disclose in the pull request description that you used an AI/LLM tool or model and how it was used.
+- You must review all AI/LLM-generated code, comments, or documentation before requesting a review.


tkuchida · 2026-10-06T22:27:52Z

+- You must disclose in the pull request description that you used an AI/LLM tool or model and how it was used.
+- You must review all AI/LLM-generated code, comments, or documentation before requesting a review.
+- You must not attribute an AI/LLM model authorship to any commit in a pull request (e.g., "Co-Authored by Claude").
+- You must answer all maintainer questions and pull request review comments yourself, without using an LLM or AI chatbot.


(This rule is unenforceable. Typically—except for things like the suspicious use of em dashes—AI is undetectable. Practically, this might just mean that someone will adjust the prompt so that the AI-generated response sounds less technical 🤷. The rule can be left here if that's the objective.)

I agree that this is unenforceable to the letter of the rule and more of a "spirit of the rule" type of thing, but I think it's valuable to include. We've had a few recent cases where it's evident that the contributor was using AI to respond to reviewers, so this gives maintainers some agency to say "hey, I think I'm talking to an LLM, that's not okay..." and point to the contributors guide. As you say, we can adjust this over time if needed.

tkuchida · 2026-10-06T22:30:02Z

+
+AI/LLM Usage Policy
+-------------------
+Contributors to OpenSim are welcome to use Artificial Intelligence (AI) or Large Language Model (LLM) coding tools to create pull requests. However, all pull requests must be created with a human in the loop. The contributor is always the author and is fully accountable for their contributions. Contributors should be confident that their submission is high quality and worth the (limited) time of the maintainers.


should -> must

nickbianco

Thanks @tkuchida! I've addressed your comments.

@nickbianco reviewed 2 files and all commit messages, and made 6 comments.
Reviewable status: 0 of 2 files reviewed, 5 unresolved discussions (waiting on adamkewley, carmichaelong, and tkuchida).

nickbianco · 2026-10-07T14:19:33Z

+
+AI/LLM Usage Policy
+-------------------
+Contributors to OpenSim are welcome to use Artificial Intelligence (AI) or Large Language Model (LLM) coding tools to create pull requests. However, all pull requests must be created with a human in the loop. The contributor is always the author and is fully accountable for their contributions. Contributors should be confident that their submission is high quality and worth the (limited) time of the maintainers.


nickbianco · 2026-10-07T14:19:34Z

+Contributors to OpenSim are welcome to use Artificial Intelligence (AI) or Large Language Model (LLM) coding tools to create pull requests. However, all pull requests must be created with a human in the loop. The contributor is always the author and is fully accountable for their contributions. Contributors should be confident that their submission is high quality and worth the (limited) time of the maintainers.
+
+- You must disclose in the pull request description that you used an AI/LLM tool or model and how it was used.
+- You must review all AI/LLM-generated code, comments, or documentation before requesting a review.


nickbianco · 2026-10-07T14:19:34Z

+- You must disclose in the pull request description that you used an AI/LLM tool or model and how it was used.
+- You must review all AI/LLM-generated code, comments, or documentation before requesting a review.
+- You must not attribute an AI/LLM model authorship to any commit in a pull request (e.g., "Co-Authored by Claude").
+- You must answer all maintainer questions and pull request review comments yourself, without using an LLM or AI chatbot.


I agree that this is unenforceable to the letter of the rule and more of a "spirit of the rule" type of thing, but I think it's valuable to include. We've had a few recent cases where it's evident that the contributor was using AI to respond to reviewers, so this gives maintainers some agency to say "hey, I think I'm talking to an LLM, that's not okay..." and point to the contributors guide. As you say, we can adjust this over time if needed.

nickbianco · 2026-10-07T14:19:35Z


 ### Looking for feedback on...

+### AI/LLM usage


nickbianco · 2026-10-07T14:19:35Z

 ### Looking for feedback on...

+### AI/LLM usage
+<!-- Please disclose any AI or LLM tools used to create this pull request.-->


tkuchida

👍 🤖

halleysfifthinc · 2026-10-07T16:45:33Z

+
+- You must disclose in the pull request description that you used an AI/LLM tool or model and how it was used.
+- You must review all AI/LLM-generated code, comments, and documentation before requesting a review.
+- You must not attribute an AI/LLM model authorship to any commit in a pull request (e.g., "Co-Authored by Claude").


I agree with the position that only humans can hold authorship (in a responsibility sense), but I think persistently attached (in the git history vs GitHub PRs/etc) indicators of LLM use are still relevant and potentially useful. The alternative I've seen and like is "Assisted-by: <tool> (model)" (e.g. "Assisted-by: Claude Code (Opus 5)")

Add AI policy to contributors guide and update PR template

272d385

adamkewley approved these changes Oct 6, 2026

View reviewed changes

nickbianco requested review from aymanhab, carmichaelong and tkuchida October 6, 2026 14:23

aymanhab approved these changes Oct 6, 2026

View reviewed changes

tkuchida reviewed Oct 6, 2026

View reviewed changes

Address review comments

980c398

nickbianco commented Oct 7, 2026

View reviewed changes

tkuchida approved these changes Oct 7, 2026

View reviewed changes

Add missed change in CONTRIBUTING.md [skip ci]

af125a9

halleysfifthinc reviewed Oct 7, 2026

View reviewed changes

Conversation

nickbianco commented Oct 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Brief summary of changes

Testing I've completed

Looking for feedback on...

CHANGELOG.md (choose one)

Uh oh!

nickbianco commented Oct 5, 2026

Uh oh!

adamkewley left a comment

Choose a reason for hiding this comment

Uh oh!

aymanhab left a comment

Choose a reason for hiding this comment

Uh oh!

tkuchida left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

nickbianco left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tkuchida left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

nickbianco commented Oct 5, 2026 •

edited

Loading