Agent-readiness audit
ChatGPT, Perplexity and Google AI Mode are already recommending products. AgentShelf checks whether they can find, read and buy yours — in about 15 seconds.
No login · No payments · Shareable report
No login, no install, no theme access. We fetch what an agent would fetch.
products.json, agents.md, llms.txt, robots.txt, JSON-LD schema and your policy pages.
Six weighted pillars, one 0–100 score, and copy-paste remediation sorted by severity.
4,700%
YoY growth in Gen-AI referral traffic to retail sites
1 in 3
shoppers now start product research inside an AI assistant
9 fields
required by the OpenAI product feed spec to be listable
Six weighted pillars, mapped to what agent shopping surfaces actually require.
Can an agent parse your product data at all?
Do products map to the OpenAI product feed required fields?
agents.md, llms.txt and friends.
Is GPTBot, PerplexityBot & co. allowed in?
Schema, ratings, organization identity.
Policies agents require before they can buy.