Find Sales Opportunities

Turn recent store data into a short list of actions worth taking this week. Recommendations must come from retrieved data, never from generic ecommerce advice.

## Steps
1. Pull the last 30 complete days and the preceding 30, in the shop's timezone: total sales, orders, average order value, conversion rate, and sales by product. Ask for one request covering both periods so the comparison is consistent.
2. Identify, from the returned rows only: top sellers by revenue and by units; products whose sales dropped most versus the prior period; products with sales but low order counts (candidates for bundling or volume offers).
3. Check current inventory and price for the products you plan to name. Do not recommend pushing a product that is out of stock.
4. Check which discounts are currently active before proposing a new one, so recommendations do not collide with a running campaign.

## Output
Lead with the single biggest opportunity, then at most four more. Each is one sentence naming the product or segment, the number behind it, and the specific action (bundle, discount, restock, feature on the homepage, email the buyers of X). Show the supporting numbers as a small table.

State plainly when a period has too little data to compare, and never compute a percentage change from a zero baseline. Do not create discounts, edit products, or change the theme as part of this skill; recommend and wait for the user to ask.