Skip to content
SEXTANTSCANSCAN A STORE

SCAN FAMILY

Agent readiness

30 CHECKS / GRADED

What an AI shopping agent can find, parse and trust when it arrives without a browser, without a session, and without patience. This is the family the engine grades, and every failing check comes back with the fix written out.

Free, no signup. Your report is a shareable link.

Sample reading. This family is graded; its arc carries a letter and a length that agree.

What it measures

  • Whether an agent can discover what the store is, from the files it publishes for exactly that purpose.
  • Whether the crawlers behind the major assistants are permitted to read the store at all.
  • Whether a product page carries the structured data an agent needs to quote a price, a variant and availability.
  • Whether returns, shipping and contact facts are stated in a form something other than a human can parse.
  • Whether titles, descriptions and images carry enough substance to answer a buyer question.

What it cannot

  • Read anything rendered only in the browser. The engine runs no JavaScript, so a store that ships its product data via client-side hydration scores badly - by design, because that is what an agent sees too.
  • Tell you whether your products are good, priced well, or in demand. It measures legibility, not merit.
  • Promise that fixing everything makes an assistant recommend you. It removes the reasons one cannot.
  • Judge more of the catalogue than it sampled. A handful of product pages and policy pages are read, not the whole store.

The 30 checks

Five categories, weighted by what actually blocks a sale. A pass earns full weight, a warn earns half, and a category that could not be measured leaves the denominator entirely rather than counting against you.

Discovery

6 CHECKS

  • llms.txt published and describes this store

    discovery.llms_txt

  • llms-full.txt expands rather than repeats llms.txt

    discovery.llms_full_txt

  • agents.md instructions for agents

    discovery.agents_md

  • Universal Commerce Protocol manifest

    discovery.ucp_manifest

  • XML sitemap is valid and populated

    discovery.sitemap

  • Product URLs discoverable from the sitemap

    discovery.sitemap_products

Crawl access

10 CHECKS

  • Served over HTTPS

    crawl.https

  • Homepage returns a successful response

    crawl.reachable

  • robots.txt is published

    crawl.robots_present

  • No blanket disallow on the whole site

    crawl.blanket_disallow

  • GPTBot is allowed

    crawl.gptbot

  • ClaudeBot is allowed

    crawl.claudebot

  • OAI-SearchBot is allowed

    crawl.oai_searchbot

  • PerplexityBot is allowed

    crawl.perplexitybot

  • Google-Extended is allowed

    crawl.google_extended

  • Homepage responds in under 3s

    crawl.response_time

Product schema

7 CHECKS

  • Product JSON-LD present

    product.jsonld

  • Offer carries a price and currency

    product.offer_price

  • Offer declares availability

    product.availability

  • SKU populated

    product.sku

  • GTIN or MPN populated

    product.gtin

  • AggregateRating present

    product.aggregate_rating

  • BreadcrumbList present

    product.breadcrumbs

Merchant data

4 CHECKS

  • Return policy is machine readable

    merchant.return_policy

  • Shipping terms are machine readable

    merchant.shipping_details

  • FAQPage schema published

    merchant.faq_schema

  • Organization schema on the homepage

    merchant.organization

Content quality

3 CHECKS

  • Product titles at least 30 characters

    content.title_length

  • Descriptions at least 500 characters

    content.description_length

  • At least 3 detectable product images

    content.images

The other sightings

Taken independently. None of them waits on another.