The challenge of details is best solved via rigorous definition of “done” (i.e. acceptance criteria). If you don’t have consensus for that, there are a million

The challenge of details is best solved via rigorous definition of “done” (i.e. acceptance criteria). If you don’t have consensus for that, there are a million ways someone else can re-define the problem and thus expose “incorrect details”. Many orgs delay this consensus, making for plenty of post code review work, but I don’t think that’s necessary bad so long as you don’t let PR comments hurt your ego.