{
    "action": "tool_call",
    "reasoning": "To identify spikes, dips, or breaks over time, I will aggregate the key metrics (orders, units sold, and revenue) by month using the order_date column.",
    "tool": "query_sql",
    "args": {
        "purpose": "Monthly time series of orders, units, and revenue to find spikes and dips.",
        "source": "raw",
        "sql": "SELECT DATE_TRUNC('month', order_date) AS order_month, COUNT(*) AS total_orders, SUM(units_sold) AS total_units, SUM(revenue_usd) AS total_revenue FROM raw.global_sports_footwear_sales_2018_2026 GROUP BY 1 ORDER BY 1;"
    }
}