top Trusting Your AI Agents to Tell You They Are "Done"

I spent three days last week watching two AI agents argue in a circle. Neither one was hallucinating. Neither one was broken. Both were being perfectly obedient to their prompts. Agent A (the coder) w