From 742c0b91fb094a0e0bdc33fff9b2bdbb84b0e7f5 Mon Sep 17 00:00:00 2001 From: Jamie Hardt Date: Sun, 3 May 2026 12:18:23 -0700 Subject: [PATCH] Update CONTRIBUTING.md Updated agents policy --- CONTRIBUTING.md | 33 ++++++++++++++++++++++++++++++++- 1 file changed, 32 insertions(+), 1 deletion(-) diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 89f2dc3..4aebcfe 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -13,4 +13,35 @@ If you discover a bug or would like better support for a feature, please do the review it as soon as I can. There's a `.devcontainer` available so you can creates commits on this project in a GitHub codespace. -Contributions made by Large Language Models are politely declined. + +## Regarding use of Agents + +`wavinfo` is an open-source project that is offered free for no commerical gain, and +is developed and maintained for educational and creative reasons. + +If you use an agent or LLM to produce code for it you are missing out on the benefits +of contributing to an open-source project, particularly community, collaboration with +other developers and designers, and being able to learn and experiment without the +burden of deadlines or worrying about business cases or profits. + +This project is supposed to be fun, do not let machines have fun for you. + +We can't prevent you from using LLMs to contribute to this project but we ask you +abide by the following eitiquette when doing so: + +* All communication with the maintainers must be written by a human in their own + voice. Never use an LLM to craft thread comments, discussion posts, issues, emails + or other correspondence with other developers or the maintainers. +* PRs must be submitted by a person. Do not allow an agent to submit its own PRs to + this project. +* Especially if you are a new contributor to this project, please submit only one PR + at a time and please restrict the subject matter of the PR to a specific unit, + module or tool. All submissions have to be reviewed and understood by the + maintainers before they can be merged. + +Obviously we can't verify if you follow all of these rules but certain telltale +traits of LLM-predicted text or code will raise a flag: lack of brevity in +descriptions or code comments, large amounts of text describing your process or +steps that add little to understanding the changes you've made, use of an +obsequious tone or being excessively accomodating, immediately doing requests +without further discussion or clarifications.