Hey all. We're doing some research on model failure paths in agentic settings and was wondering if people have seen this in the real world. Have you noticed any shady differences when updating model versions or changing providers? How did you work around them? Like hallucinations or models doing something else.