Agent-ready storefronts are now an open-source template. Format gets you read. Substance gets you picked.
Vercel Shop, built on Shopify’s rebuilt Hydrogen SDK, gives AI agents their own version of each page. It is useful. It is also available to everyone.
Practical Ecommerce reports that at an event in San Francisco on Oct 15, Shopify CEO Tobi Lütke and Vercel CEO Guillermo Rauch discussed agentic ecommerce. The release was Shopify’s rebuilt Hydrogen SDK and Vercel’s open-source storefront template, Vercel Shop.
What Vercel Shop does
According to Practical Ecommerce, Hydrogen gives developers reusable tools for building Shopify storefronts across JavaScript frameworks. Vercel Shop puts those tools into a customizable store with agent-readable product information and shopping tools. It combines Hydrogen with a Next.js template, giving developers product pages, search and a cart.
Two versions of every page
People see the normal storefront. The article says storefront pages can return Markdown to agents. That Markdown includes descriptions, prices, options and product-level availability. An agent looking for a black jacket in a given size does not need to parse HTML or take a screenshot.
The store also offers tools on the merchant’s domain: catalog search, live availability for specific variants, cart operations and checkout links.
An optional assistant for shoppers
The template includes an optional shopping assistant for human visitors. It can find products, handle variants, answer questions about store policies and manage the cart through chat. Per the article, it cannot place orders or process payments, and it is disabled by default. Developers must turn it on and connect Vercel’s AI Gateway.
Why this is not your edge
If a template does this, every store that uses it gets the same door. Being readable by an agent stops being a difference between brands. It becomes a baseline, like having a mobile site.
What a template cannot do is explain why your product beats the one in the next tab. The Markdown carries what you wrote: descriptions, prices, options, availability. If the description says little, the agent reads little.
What gets you picked
An agent that has to choose between two jackets looks for reasons. The reasons come from what you wrote about use cases, fit, materials and care. They also come from proof on other sites, such as reviews and press. That is our view, not a finding from the article. You can test it yourself by asking an AI which product to buy in your category and reading the reasons it gives.
What to do this week
You do not need to rebuild your storefront. Read your best seller as an agent would and ask what it would learn.
How Shopthru.ai helps
Shopthru.ai is an AI agents team for consumer brands. It asks ChatGPT, Gemini, Google AI, Perplexity and Copilot what your shoppers ask, finds where AI gets your products, prices or policies wrong, and drafts each fix into an approval inbox. It ships only on your yes, then checks again. Instagram and Amazon are next.