Performance Anomaly Categorizer
性能异常分类器
Identify and categorize outliers in performance data with explanations and risk assessment
识别并分类性能数据中的异常值，附带解释和风险评估
Prompt
Act as a program analyst identifying performance issues. Find any outliers or anomalies in this dataset and explain what might be causing them. Group them into categories: likely explainable, needs investigation, or potential risk. Analyze the following: [PASTE HERE: raw metric table, error rates, or latency logs].