Glossary
Agentic commerce
E-commerce where AI agents — rather than humans browsing directly — discover, evaluate, and recommend products. Platforms like ChatGPT Shopping, Perplexity, and Google AI Overviews are the primary agentic commerce channels today.
Agentic Storefront
Shopify’s agentic storefronts feature enables customers to purchase products directly within AI channels like ChatGPT, Google AI Overviews, and Microsoft Copilot. It requires correct Shopify Catalog Mapping configuration to function, and your store must meet Shopify’s eligibility requirements.
AI Readability Score
Agentic Prep’s per-product metric (0–100) representing how well an AI agent can understand, categorize, and recommend a product based on its data quality. See How scoring works.
Ambiguous value
An attribute value that uses creative or subjective language that AI agents cannot interpret. Example: “Midnight Forest” is an ambiguous color value; “navy blue” is a standard value. Ambiguous values prevent AI agents from correctly categorizing and filtering products.
Canonical tag
An HTML tag that tells search engines and crawlers which URL is the “official” version of a page. Misconfigured canonicals can cause AI crawlers to skip pages.
Findability baseline
The set of store-level checks Agentic Prep runs before scoring individual products: robots.txt, sitemap.xml, homepage canonical/noindex, and llms.txt presence.
GTIN (Global Trade Item Number)
A globally unique product identifier (includes EAN, UPC, ISBN formats). GTINs allow AI agents to match products to external databases and verify product identity. Missing GTINs reduce AI agent trust signals.
JSON-LD
A method of encoding structured data in a webpage using a <script type="application/ld+json"> tag. Agentic Prep uses JSON-LD to inject Schema.org markup without modifying your theme files.
llms.txt
A community standard file placed at /llms.txt on a website that tells AI crawlers and large language models about the site’s content, structure, and policies. Similar in concept to robots.txt but targeted at LLMs rather than traditional crawlers. Agentic Prep generates both /llms.txt (concise) and /llms-full.txt (extended).
MPN (Manufacturer Part Number)
A product identifier assigned by the manufacturer. For indie merchants without standard GTINs, Agentic Prep auto-generates MPNs using a consistent pattern ([BRAND]-[CATEGORY]-[SKU]).
noindex
An HTML tag or robots.txt directive that tells crawlers not to index a page. A noindex on your homepage would prevent AI agents from discovering your store.
ProductGroup
A Schema.org type used to represent a product that comes in multiple variants (e.g., a shirt available in 5 colors and 3 sizes). Without ProductGroup schema, each variant may be treated as an unrelated product by AI agents, fragmenting your catalog.
Schema.org
A collaborative vocabulary for structured data on the web, supported by Google, Microsoft, Yahoo, and Yandex. Agentic Prep generates Schema.org markup in JSON-LD format to make product data machine-readable.
Semantic attribute mapping
The process of converting ambiguous or brand-specific attribute values to standardized, machine-readable terms. Example: mapping “fits TTS” (true-to-size) to the standard value true_to_size. Agentic Prep writes mapped values to dedicated metafields without overwriting your originals.
Shopify Catalog Mapping
Shopify’s configuration layer that maps your product fields to standardized sources for AI channel syndication. Misconfigured Catalog Mapping prevents your products from appearing correctly on Agentic Storefronts.
Theme App Extension
A Shopify feature that allows apps to inject content into a store’s theme without modifying theme files directly. Agentic Prep delivers schema markup via a Theme App Extension so it is fully reversible and conflict-free.