I need to identify columns with missing values and calculate the percentage of missing data for each. Let me query the column metadata to get null counts across all tables. ```json { "action": "tool_call", "reasoning": "The analytics.column_overview metadata table contains null_count for every column across all tables. I'll query it to find columns with missing values and calculate the percentage missing based on each table's row count.", "tool": "query_sql", "args": { "sql": "SELECT co.table_name, co.column_name, co.null_count, t.row_count, ROUND(100.0 * co.null_count / t.row_count, 2) AS pct_missing FROM analytics.column_overview co JOIN analytics.table_overview t ON co.table_name = t.table_name WHERE co.null_count > 0 ORDER BY pct_missing DESC, co.table_name, co.column_name", "purpose": "Find all columns with missing values and calculate the percentage missing for each" } } ```