My idea was that, if I see that your project is growing 10k LOC per week and you're the only developer working on it, it's most likely vibe-coded.
I analyzed some open-source projects, but unfortunately it turns out not to be so clear cut. It's relatively easy to estimate the growth rate of a project, but figuring out how much time developers worked on it is very error prone, which results in both false positives and false negatives.
I wrote a post about it (https://pscanf.com/s/352/) if you're interested in the details.
And then look through the commits -- were they only adding new features, or did the author(s) put effort into improvements on engineering fundamentals (benchmarking, testing, documentation, etc)?