Your AI agent sees every number you see; Products and Support reports

• Analytics → Reports → Products. Units sold, sales, products that sold, and the share of units refunded or returned, against the period before. One table has every product: units, sales, average price, units per day, refunds, return rate, gross margin and rating, plus your best-selling variants. If you track stock, you also see what is out of stock now and what runs out within 14 days.
• Analytics → Reports → Support. Tickets opened and closed, how many are waiting for a reply, tickets per 100 orders, first reply time, time to close, replies by the AI assistant and by your team, and why customers write in.
• Automatic "we got your message" emails no longer count as a reply. Before, they made first reply times look instant and hid tickets from Awaiting reply.

New for agents

• dtc_store_snapshot. Every area in one call: sales, profit, payments, ads, traffic, subscriptions, products, delivery, support and chargebacks. Each number comes with the period before and the change.
• dtc_metrics_catalog and dtc_metrics_query. More than 40 reports, the same as your Analytics screens and computed by the same code: full daily series and every table row (per campaign, ad account, product, carrier, ticket reason and more).
• dtc_metrics_export. Raw rows for the agent's own analysis: orders, order lines, refunds, ad results, subscriptions and their changes, parcels, support tickets and chargebacks. Customers appear as ids, never names, emails or addresses.
• Agents learn all of this from platformdtc.com/SKILL.md. See Store numbers for agents.