Continue? Y/N: A 60-second game about AI agent permission fatigue

This is a browser game, not a tool, but it earns a brief because the thing it is satirizing — the endless confirm-before-acting loop in agentic AI systems — is a genuine UX and product design problem that every team building agent workflows is wrestling with right now. Sixty seconds of rapid-fire Y/N permission prompts communicates the problem more vividly than a blog post would. It is the kind of thing you send to the team Slack to frame a conversation about where to draw the boundary between autonomous action and human-in-the-loop checkpoints. Reservation: it is a one-joke demo. You will get the point in under a minute and never open it again. But if it prompts a real conversation about agent permission design in your product, that minute is well spent. The underlying question — how much autonomy is the right amount — is one of the more important UX problems in the current wave of agent-based products. -> Best for: SaaS team of 2-5 designing agentic product workflows