A firsthand look at debugging persistent AI agent memory, from multilingual retrieval failures and outdated instructions to a temporal search bug.