How to deal with fully LLM-generated merge requests

[Luca Beltrame]

Il 10 settembre 2026 20:33:09 CEST, Nate Graham <nate at kde.org> ha scritto:

Hello Nate,


>Perhaps this is a good time to propose a specific set of LLM usage guidelines.
>
[snip]

Overall I think it's in a good direction. Some minor points (my opinion, feel free to criticize).

>3. Don't submit "vibe-coded" changes you don't understand and couldn't make yourself.

I would be explicit about what "vibe coded" is, and add the term in brackets. While I use LLMs in my day to day work, it took a while for me to understand what the term meant.

>4. Don't disclose LLM usage as a way of trying to excuse the potential errors or poor quality of a contribution.

"If there are potential errors or the contribution is of poor quando, don't disclose LLM usage as an attempt to excuse for them"?


>### The golden rule
>**Don't be lazy.**
>
>If anyone can detect that you used an LLM, you were being too lazy.

I think it's a bit too strong here. Some will succeed. Some will be asked to revise after the initial submission.

Also, I would like to ensure, on part of who has to review potential LLM-assisted contributions, that there is good faith involved.
This is to avoid unfortunate comments like in a recent kdepim PR.