example-walled.com

Whether your product's structured data is in the page's initial HTML or only appears after JavaScript runs. ChatGPT, Claude, and Perplexity read the initial HTML — anything that needs JavaScript is invisible to them.

Whether your structured data is valid: it parses cleanly, required fields are present, values are the right type, and nothing contradicts the visible page.

How complete your structured data is beyond the basics — product ID (GTIN), manufacturer part number (MPN), ratings, price validity, and return and shipping policy.

Structured data across the rest of the site, not just the product page — your organization details, site search, breadcrumbs, and FAQ markup.

How much real product text remains once JavaScript has run. Copy hidden inside tabs or accordions can be invisible to AI agents.

Whether the page a shopper sees and your structured data agree on price, availability, brand, and title. Mismatches read as a trust problem to an agent.

How usable your product images are to AI: enough images, descriptive alt text, image structured data, and responsive sizes. Agents increasingly read images, not just text.

Signals about how recently the page was updated. Informational for now (not scored) until a reliable signal lands.

Whether the product page links out to its category, related products, and brand pages — the trail an agent follows to understand your catalog.

Whether the site publishes an llms.txt file — an emerging convention for pointing AI agents to your key pages. Low weight while adoption is early.

Whether the site publishes an /agents.md file — an emerging convention, alongside llms.txt, for pointing AI agents to a site's real agent endpoint. Informational for now while the standard is nascent.

Whether the storefront exposes a reachable Model Context Protocol (MCP) endpoint — the emerging way for AI agents to query a catalog and act on it directly. Informational for now while the standard is nascent.

Whether product titles are structured and descriptive — brand, product type, a key attribute, and the variant — rather than a bare product name.

Whether descriptions are rich with real attributes and answer the questions a shopper actually asks, rather than marketing fluff.

Whether the page has the question-and-answer pairs, usage scenarios, and compatibility notes an AI agent can quote back to a shopper.

Whether the core product identifiers — GTIN, MPN, brand, and model number — are present and actually valid.

Whether availability is specific — exact stock, handling time, and restock dates — rather than a bare in-stock / out-of-stock flag.