How agent-ready is your catalogue?
A free self-assessment that scores how discoverable your store is to AI shopping agents and answer engines, and what to fix first. It is a guide, not a guarantee.
01 / What it checks
The signals AI agents actually read
Shoppers increasingly start with an assistant, not a search box. These are the things that decide whether an agent can find, understand and act on your products.
Structured product data
Valid Product schema, filled attributes and a clean taxonomy are what a machine reads first. Free-text descriptions are not enough.
Machine-readable feeds
A current, accepted product feed and search that handles natural language decide whether an agent can find the right product at all.
Honest AI signals
Disclosing AI assistants, marking AI-generated media and setting a deliberate crawler-access policy keep you on the right side of the rules.
Agent feeds
Open feeds such as ACP and MCP let assistants act on your catalogue, not just read it. This is the newest and least adopted part of the check.
02 / Self-assessment
Score your catalogue
Answer ten honest questions about your store. We total your answers and place you in one of three bands, with plain guidance on what to do next. Nothing you enter is sent or stored.
0 of 10 answered
01 / Structured product data
Do your product pages carry valid Product structured data (schema.org) with price, availability and identifiers?
02 / Clean attributes
Are core attributes (brand, colour, size, material, GTIN/MPN) filled in consistently rather than buried in free-text descriptions?
03 / Taxonomy
Is your category and tag structure consistent, deduplicated and mapped to a standard taxonomy (such as Google's product taxonomy)?
04 / Machine-readable feed
Do you publish a current, complete product feed (for example a Merchant Center or equivalent feed) without large rejection counts?
05 / On-site search
When a shopper types a vague or natural-language query, does your site search return relevant results rather than nothing?
06 / Returns and fit data
Do your pages expose structured sizing, fit or compatibility data (size charts, dimensions, what-fits-what) that a machine can read?
07 / AI interaction disclosure
If shoppers interact with an AI assistant on your site, are they clearly told they are talking to AI, with a route to a human?
08 / Synthetic-media marking
Where you use AI-generated images or video, is that media clearly marked as AI-generated?
09 / Agent access
Do your robots and access rules let reputable AI crawlers and shopping agents read your catalogue (rather than blocking them by default)?
10 / Agent feeds (ACP/MCP)
Can an AI agent discover and act on your catalogue through an open feed such as the Agentic Commerce Protocol (ACP) or Model Context Protocol (MCP)?
Answer every question to see your band.
03 / Honest note
What this is, and what it is not
This check is a guide to help you spot gaps, not a technical audit and not a guarantee of how any particular agent will treat your store. The standards behind it, from structured data to agent feeds like ACP and MCP, are young and still changing, so treat AI readiness as ongoing work rather than a box to tick once.
Most of the fixes that raise your score, such as cleaner attributes, a healthier product feed and search that understands plain language, help ordinary shoppers and search engines too. That is the work worth doing first, whatever AI agents do next.
Build on a catalogue agents can read
Sellarix keeps structured data, feeds and agent access in one place. Join the waitlist, or talk to us about your store.