But it makes you feel like Cassandra when you keep warning people about the same problems, and keep running into them months and years down the line.

You can only post the Surprised Pikachu meme so many times before it loses its luster.

Or incompetent. If you knew this was going to break, why did you approve it? Your only defense is "there was a lot pressure to get into the release xyz". There's not much sympathy for that defense. The animal spirits that thought the broken feature were the most important thing ever are long gone, and frustrations about the new outage caused by the previously most important feature ever have taken over.

Don't approve it! That's actually a thing you can do.

Often it's more subtle than just "this is clearly going to take down production in X way at some point." The issue is more like feeling that the logic is too entangled and it's going to be hard to maintain later on, or that a library should or should not have been used, or something done with threads should have been async.

So yeah, not as cut and dried as "I said it would happen and it did" but more like "I had a feeling this was going to turn out to be a pain and sure enough here I am reviewing code that represents that pain."