If # of commits in PR ≥ 20, assume accidental PR against wrong repo

[cvrebert]

(Or botched rebase, or etc...)
Refs twbs/bootstrap#14351

[hnrch02]

Could be problematic because of for example twbs/bootstrap#14951.

[cvrebert]

That PR is a statistical outlier, but yeah, the limit can be debated and twerked.
And/or maybe only enforce it on the initial opening of the PR.

[cvrebert]

Maybe also put in a heuristic about the commits having N different authors and the requester not authoring a majority of the commits.

[hnrch02]

That sounds good. Looking forward to the limit being twerked! 😄