See your store become readable by machines
Explore a sample Shopify product before and after AgentReady. See what machines are missing, review the proposed structured data, and inspect the evidence AgentReady checks after publishing. Everything below is one fictional store, walked through the real product loop.
1Human view / Machine view
The same product, seen twice
A shopper sees the page on the left of the split. A machine reads the structured facts on the right, each one carrying who published it and when it was last verified. Sample data for a fictional store.
Two renderings of the same fictional product page: the human view a shopper sees, and the machine view of structured facts AgentReady publishes and verifies. All values are sample data.
Human view
Northbound
Trail Runner 2
$140.00
In stock
A lightweight trail shoe with a grippy outsole and a roomy toe box, built for long days on mixed terrain.
Add to cartFree US shipping over $75 · 30-day returns
Machine view
- @type
- Productours
- name
- Trail Runner 2ours
- brand
- Northboundours
- sku
- NB-TR2-9ours
- gtin13
- 0791234567890ours
- variants
- 4 sizes, one product groupours
- price
- 140.00ours
- priceCurrency
- USDours
- availability
- InStockours
- shippingDetails
- Free over 75.00 USD, USours
- returnPolicy
- 30-day returnsours
- seller
- Northbound Outfittersours
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Trail Runner 2",
"brand": { "@type": "Brand", "name": "Northbound" },
"sku": "NB-TR2-9",
"gtin13": "0791234567890",
"offers": {
"@type": "Offer",
"price": "140.00",
"priceCurrency": "USD",
"availability": "https://schema.org/InStock"
}
}AgentReady fetched the live page and confirmed its own output is present and valid. Sample data, dated at the check.
last verified: Jul 31, 2026 · sample data
2Before AgentReady
What a machine reads on the sample page today
Presentation markup with no machine-readable facts. An agent has to guess the price, brand, and availability.
<div class="product"> <h1>Trail Runner 2</h1> <p>$140.00</p> <button>Add to cart</button> </div>
3The audit
What's missing, and what disagrees
The audit reads the sample storefront the way a crawler does and reports two kinds of problems: facts machines cannot find, and facts stated twice with different answers.
The page never states who makes the product, so an assistant cannot attribute it.
In stock or not? The button says Add to cart; machines do not infer from buttons.
The store has a 30-day policy page, but nothing connects it to this product.
A leftover theme snippet still announces $150.00 while the page shows $140.00. Two answers to one question is worse than none.
4The proposals
AgentReady prepares the fix, from facts the store already holds
Every proposal names its source. Nothing is invented: if the store does not hold a fact, AgentReady asks instead of guessing.
From the catalog entry Shopify already holds
One structured block stating name, brand, current price, currency, and stock state, generated from the store's own data, never invented.
From the store's existing policy page
The 30-day return window the store already publishes becomes a machine-readable fact on the product itself.
From the live catalog price
The stale $150.00 snippet is flagged; the proposal carries the one true price so machines stop seeing two answers.
5Your approval
Nothing publishes until the merchant says so
Each proposal is reviewed with explicit Approve and Discard, and a preview of the exact content. Existing values stay until you choose to replace them. In this sample, the merchant approves all three.
6After AgentReady
What a machine reads once the approved output is live
Valid Product JSON-LD with brand, price, availability, and the connected return policy, published through AgentReady-owned theme blocks that carry their own provenance.
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Trail Runner 2",
"brand": { "@type": "Brand", "name": "Northbound" },
"offers": {
"@type": "Offer",
"price": "140.00",
"priceCurrency": "USD",
"availability": "https://schema.org/InStock"
},
"hasMerchantReturnPolicy": {
"@type": "MerchantReturnPolicy",
"merchantReturnDays": 30
}
}7The publish state
Applied is not the same as verified
After a publish, AgentReady says exactly which of these two states the store is in, and never dresses the first up as the second.
Your approved changes are applied. We are checking the live storefront now.
Your store is live and verified: the check found AgentReady's own structured data on the public page.
8The verification
What the live check actually proves
The check fetches the storefront the way an AI crawler does. A password page or blocked robots file is reported as unknown, never guessed.
It looks for valid structured data on the live page, not in a preview or a cache.
Provenance matters: theme or app markup that AgentReady did not publish never counts as AgentReady output.
Every block carries its generated-at timestamp, so a stale copy is visible instead of silently trusted.
9Ongoing
The loop keeps running after the first publish
- Catalog edits re-publish through the same approval gate you already used.
- Drift detection compares the live page with your store and flags mismatches.
- Uninstalling removes AgentReady's output and leaves your theme clean.
Run the same read on your own store
The free AI-readiness checker fetches your storefront the way an AI agent does and scores what it finds, in seconds, no install required.