Extended Abstract “Reinforcement Learning As End-User Trigger-Action Programming” accepted at IMLW@AAAI-22