It would have been easy to run a few ai agents to review the code and find these issues as well and architect it clearly