store-sales-dashboard

Run the dashboard subcommand with a date and optional store or region filter, for example: python store_sales_dashboard.py dashboard --date 2026-09-03 --store store-01. The script returns JSON metrics including gross sales, net sales, customer count, and average basket size.

Use the budget-variance subcommand with a period (daily, weekly, monthly) and date. Budget targets are loaded from a budgets.yaml file defining monthly, weekly, and daily budgets per store, and the report flags stores exceeding or falling below thresholds.

The skill expects a sales transaction database configured through the SALES_DB_URL environment variable, defaulting to a PostgreSQL connection at postgresql://localhost:5432/sales. Store identifiers are set via SALES_STORE_IDS.

No, the dashboard has a data latency of 5-15 minutes behind POS systems, so it is not suitable for real-time monitoring. Data freshness should be labeled on dashboards, and same-day returns can distort net sales figures.

Sequential day comparisons during holiday periods produce misleading trends because shopping patterns shift with the calendar. Year-over-year comparisons against the same calendar day give more meaningful results for holiday analysis.