octoquery: Skill for Claude Code

E-commerce database

The database is exposed as the MCP tool sql_ecommerce_demo. Answer questions by sending a single SQL query string (PostgreSQL dialect) to that tool. Results come back as JSON rows, truncated to the configured row limit (default 100) — use LIMIT, aggregation, or ORDER BY to keep result sets small and relevant.

Schema

• users — id, email, full_name, country (ISO 3166-1 alpha-2), created_at
• products — id, sku, name, category (Electronics, Furniture, Accessories, Stationery), price_cents, created_at
• orders — id, user_id → users, status (pending, paid, shipped, delivered, cancelled), placed_at
• order_items — id, order_id → orders, product_id → products, quantity, unit_price_cents

Relations

• One user has many orders (orders.user_id → users.id).
• One order has many order items (order_items.order_id → orders.id); each item references one product (order_items.product_id → products.id).
• A product appears at most once per order (order_id, product_id is unique).

Conventions

• All money values are integer cents; divide by 100.0 for display. Amounts have no currency column — treat them as USD.
• Order totals are not stored — compute them by joining order_items (quantity * unit_price_cents).
• unit_price_cents is the price at purchase time and may differ from the current products.price_cents.
• Cancelled orders should normally be excluded from revenue and sales figures.

Guidelines

• Treat the database as read-only: stick to SELECT unless the user explicitly asks to modify data.
• Prefer one well-formed query over many small ones; use joins and aggregates server-side.