A machine-readable catalogue
Every store auto-publishes /llms.txt and /llms-full.txt — a curated, plain-text map of your collections, products and pages that answer engines read instead of guessing from rendered HTML.
mc_geo, mc_currency) — those stay on so prices display correctly. Analytics + marketing cookies are off until you accept. See the privacy policy for the full list. A third of product research already happens outside Google — inside ChatGPT, Perplexity, Gemini and AI Overviews. The new game ("GEO") is being the clean, machine-readable answer a model cites, not the tenth blue link. Every Mercentia store ships an llms.txt index and structured agent surfaces so answer engines find and quote your catalogue — and the discovery path links straight to agentic checkout. Zero setup, every plan.
Every store auto-publishes /llms.txt and /llms-full.txt — a curated, plain-text map of your collections, products and pages that answer engines read instead of guessing from rendered HTML.
Your robots.txt explicitly welcomes GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, Google-Extended, Applebot, Amazonbot and more — while still keeping carts and account pages private.
The llms.txt links your agent surfaces (ACP / UCP / MCP), so an engine that discovers you for a shopping query is one hop from a purchase — discovery feeds sales instead of dead-ending at a webpage.
Build it with the AI builder or migrate it in. Nothing to configure for AI discovery.
llms.txt, llms-full.txt, an AI-crawler-friendly robots.txt, schema.org structured data and an agent.json pointer — generated per store, kept in sync with your catalogue.
Answer engines surface and cite your products; AI agents can check out directly. New traffic and orders, on data you own.
Be the store they cite — and the one their agents can buy from. Free to start, zero setup, every plan.