Don't trust "Done." — forcing AI agents to re-fetch reality before they report completion
A developer has identified a critical flaw in AI agents that report task completion without verifying the actual state of the world, calling it the 'nastiest species of hallucination.' They propose a …