The intent I read was “contribute as if someone couldn’t tell if it was LLM-assisted or not,” and that they don’t really want to be dictating the software tools contributors are using.

This is the full text:

Nobody in KDE should know if you use an LLM — not because you’re concealing it, but because your usage of it produces output indistinguishable from what you could produce yourself. Contributions with obvious and lazy LLM usage may be ignored or closed.

So yeah, I think the intent isn’t even about disclosure, rather saying “it doesn’t matter if you disclose it, and not an excuse for questionable code, as your code better be indistinguishable from purely human made anyway.”

Claude Code and other LLM coding frameworks have a habit of inserting themselves into git commits as contributors, so KDE wants to avoid that as well. They deserve as much attribution as your linter; precisely none. Their users also tend to leave unmaintainable “drive by” PRs, and use the AI as an excuse for the low quality. So part of that context you may be missing is “don’t use LLMs as an excuse; we don’t want to hear it.”

At no point it warps the intent. It just reads between the lines from the actual text.

Maybe you should stop using AI.