I need to identify all numeric columns in the dataset and compute their distribution statistics in a single efficient query. Let me start by inspecting the columns of the main table to see which ones are numeric: ```json { "action": "tool_call", "reasoning": "I need to identify all numeric columns in raw.all_ai_models to compute their distributions. The card shows many measures but I need the exact column names and types to write an efficient aggregation query.", "tool": "inspect_columns", "args": { "table": "raw.all_ai_models" } } ```