How do I uncover hidden constraints before it's too late?

You can run technical discovery now, and still the buried constraints keep surfacing too late, after decisions are made and the fix is expensive. That is the hardest part of discovery, because the constraints that hurt most are exactly the ones nobody thinks to mention until you trip over them. The reason they stay buried is that the people who know them do not know they matter. To them it is just how things are. So they never say it, and you never ask, because you do not know it exists. The skill is not asking more questions. It is asking the kind that surface assumptions people do not know they are holding. Start keeping a running list of the constraints that bit you late on past projects, and turn each into a question you ask early next time. What has to keep working, what cannot change, what has burned us before? Discovery gets sharper the same way anything does, by learning from the exact things that got past you last time. You cannot catch every hidden constraint. You can stop getting caught by the same ones twice.