Nxora · Shopify apps

AI Catalog Readiness

Find out why your products aren't reaching AI assistants.

This app checks every product against the criteria Shopify publishes for its Catalog and Agentic Storefronts pipeline — the feed that puts products in front of assistants like ChatGPT, Copilot and Gemini. It scores your catalog 0–100, and ranks what to fix first by how much revenue sits behind each product.

Every finding names the field, says what is wrong, and explains why an AI assistant cares.

Where this sits next to Shopify's own tool

Shopify ships a listing quality indicator inside Agentic Storefronts, it is free, and it is already in your admin. Use it.

Shopify scores the products that already made it into Catalog. This tells you about the ones that did not. That is the boundary, and we are not going to claim more than it. This is the step before their tool, not a replacement for it — you run this to find the products that never got in, fix them, and then their indicator has something to score.

What the report gives you

• A score, 0–100, and the formula behind it. Critical findings −25, major −12, minor −4, averaged across products. Published, so you can recompute it from the findings yourself. No black-box number.
• A "fix these first" list, ranked by revenue at stake. Severity multiplied by commercial exposure, so you fix the 12 products that matter, not all 4,000. This is what read_orders is for — see below.
• Findings per product, each with a why. The field, what is wrong with it, and the reason an assistant treats it as a disqualifier. Missing barcodes, unnamed options still called "Option 1", empty categories, incomplete variants, thin descriptions.
• A catalog-wide rollup. Where one fix repairs hundreds of products at once. Seeing 380 × identifier.barcode_missing in a single row tells you the size of the win before you start.
• Change tracking between scans. When a product's set of defects actually changes, you hear about it. An ordinary price edit that breaks nothing is not an alert.

Why the app asks to read orders

read_orders on a free diagnostic is a real ask, so here is the answer before you hit the permission screen. It is used for exactly one thing: counting units sold per product over the last 60 days, so the fix-list can be ordered by commercial impact instead of arbitrarily.

The query asks for no customer field at all — no name, no email, no address, no payment detail. It selects only lineItems { quantity, product { id } }, and the result is reduced to one integer per product before anything is stored. No order record is kept, and no customer personal data is collected anywhere in the app. The privacy policy sets out every table we store and what is in it.

Pricing

Billed through Shopify. The Pro tier's AI review reads the top of your worklist — up to 25 products per review — and shows its suggestions beside your current text. It never writes them, because the app has no write access to write them with.

Questions

Will it change anything in my store?

No. It cannot. The app holds read-only scopes and requests no write permission of any kind.

Does it collect my customers' data?

No customer personal data is collected at all. Details, table by table, are in the privacy policy.

What happens when I uninstall?

The app's access token is deleted and it loses all ability to read your store at that moment. Your stored scan reports contain scores, finding codes, product ids and titles — email us and we will delete them. See the retention section of the privacy policy.

Do I still need Shopify's own listing indicator?

Yes, and you should use it. It scores what is in Catalog; this finds what is not.

Contact

support@getnxora.com. Real address, monitored by the people who build the app.