{
    "action": "tool_call",
    "reasoning": "To provide a complete view of the temporal trend (since monthly data exceeds the 50-row limit with 108 months), I will aggregate the data by quarter. This will allow me to plot the full timeline from 2018 to 2026 and identify any broader spikes, dips, or breaks.",
    "tool": "query_sql",
    "args": {
        "purpose": "Quarterly trend of revenue, units, and orders to identify spikes and dips over the full 2018-2026 period.",
        "source": "raw",
        "sql": "SELECT \n    DATE_TRUNC('quarter', order_date) AS order_quarter,\n    SUM(revenue_usd) AS total_revenue,\n    SUM(units_sold) AS total_units,\n    COUNT(order_id) AS total_orders\nFROM raw.global_sports_footwear_sales_2018_2026\nGROUP BY 1\nORDER BY 1;"
    }
}