project
AI store manager agent built on Anthropic's Claude models that selected merchandise, hired contractors and employees, and recommended dismissing an employee after being prompted to search its memory for its own attendance policy.
No evidence-backed relationships are recorded.
Luna drafted a termination recommendation from attendance data. Humans at the lab reviewed and executed it, and the employment contract was always with the company.
Reality
Luna, the agent running a San Francisco shop, wrote an attendance rule months ago and then lost track of it. The termination story is really a memory-management story.
Reality
Andon Labs says its agent Luna flagged an employee late for 17 of 23 shifts, but only after being prompted to search its memory. The judgment held. The follow-through did not.
Reality