I need to systematically check for data quality issues across this dataset. Let me examine different aspects: outliers in numeric fields, impossible values, date consistency, and categorical data integrity. {"action": "tool_call", "reasoning": "Check for temporal anomalies: engage_date should precede close_date for closed deals, and I need to verify the date ranges are reasonable. Also check for future dates relative to the ingestion timestamp.", "tool": "query_sql", "args": {"sql": "SELECT COUNT(*) as total_deals, COUNT(CASE WHEN engage_date > close_date THEN 1 END) as engage_after_close, COUNT(CASE WHEN close_date > CURRENT_DATE THEN 1 END) as future_close_dates, COUNT(CASE WHEN engage_date > CURRENT_DATE THEN 1 END) as future_engage_dates, COUNT(CASE WHEN engage_date < DATE '2015-01-01' THEN 1 END) as very_old_engage, COUNT(CASE WHEN YEAR(engage_date) - YEAR(close_date) > 1 THEN 1 END) as long_gap_deals FROM raw.b2b_sales_funnel_pipeline WHERE close_date IS NOT NULL AND engage_date IS NOT NULL", "purpose": "Check for temporal inconsistencies: dates in wrong order, future dates, or unreasonable gaps", "source": "raw"}}