The agent asks you questions until you have clarity on what to do.
Experimenting with this idea through AntiAgent. I'd love Hacker News to challenge this idea.
Show HN: What if AI did the prompting – and humans did the thinking?
antiagent.site