See what AI agents do on your store or website.
AI Agent Snapshot shows who's visiting, what they're likely looking for, which pages and commerce endpoints they access, and how they move through your store. Real detection. Not estimates.
Launching as a Shopify app. Built first for the stores where AI commerce is happening, coming to websites next.
AI agents aren't just visiting. They're deciding.
AI agents are starting to interact with stores. They search, retrieve, inspect products, check policies and access commerce capabilities. Most ecommerce analytics were built to understand humans. AI Agent Snapshot shows what AI agents do instead.
If you need the one-line comparison: think Google Analytics, but for AI agents. That's the mechanism. What matters is what it reveals.
Built for founders.
Minimal gear for people building companies, products and ideas.
Illustrative activity based on real traffic patterns on our own store, founder.atomfoundry.dev.
Just a signal. And an inbox.
No dashboard to babysit. No reports to interpret. When an AI agent reads your store, you see it.
One small icon lives quietly in your admin, visible only to you and your team, no one else. Click it, and a few dials fan out at the edge of the screen, like an instrument cluster. Click again, they are gone. Click through and you can open the full view for a complete history. The rest happens on its own: the moment an AI agent reads your store, an email goes out that says exactly what happened.
At 14:32, Perplexity-User visited /products/founder-trucker-cap, /collections/apparel, and /policies/refund-policy.
This is the first visit from this agent in 6 days.
No action needed, just keeping you posted.
Full history, alert recipients and how often you hear from us are all configurable in Settings.
Not just traffic. A record of what happened.
Which agents keep coming back, how often, and which pages they're pulling from your store.
| Agent | Pages | Frequency | Last seen |
|---|---|---|---|
| Claude-Web | 4 | 3× | 2m |
| ChatGPT-User | 7 | 12× | 8m |
| Perplexity-User | 3 | 2× | 21m |
Illustrative 30-day trend based on real traffic patterns on our own store.
AI is interested in this product. AI keeps reading our return policy. Signals like these can shape your entire AI commerce strategy.
It's not "ChatGPT visited 17 times." It's why.
Every AI agent visit has a likely intent behind it. Knowing what an agent was trying to do turns a raw visit count into something you can act on.
ChatGPT came back 17 times to check these three products and the shipping policy.
Not a list of pages. A path.
The order an agent moves through your store tells you more than which pages it hit. It tells you what it was building toward, and where it stopped.
Classic ecommerce analytics doesn't show this second kind of journey at all. It's not a page view, it's a protocol handshake, an agent checking what a store lets it do before it does anything.
What gets looked at. What gets skipped.
An agent that requests a collection page but only follows up on 2 of 20 products is telling you something: those products were surfaced, the rest weren't picked up on. This is inferred from request patterns, not measured dwell time, and it's still developing.
Not "how many visited." Where do they stop.
A funnel view of the same visits shows where agent activity falls off, and by how much at each step.
From AI reads your store to AI interacts with it.
As Shopify stores expose more of their functionality to AI agents as tools, not just pages, agents move from reading a store to acting on it: searching, filtering, adding to cart, checking out. This is where AI Agent Snapshot is going next.
Once agent activity can be reliably linked to a commerce event, we can say something no traffic dashboard can: these AI agents did not just visit your store. They tried to buy.
Not just an agent name. What kind, and what it's doing.
A user agent string tells you a request came from ChatGPT. Classification tells you whether it's a shopping agent comparing options, a crawler indexing pages, or something you haven't seen before.
| Agent | Type | Behavior |
|---|---|---|
| Claude-Web | Research agent | Researching |
| ChatGPT-User | Shopping agent | Comparing |
| PerplexityBot | Search agent | Crawling |
| Unknown | Unknown | Testing |
Detect. Classify. Understand. Then control.
You already get detect (who came), classify (what it is and does), and understand (why it came, what it's looking for). Control, what you allow an agent to do, is the next layer.
Your store, or the whole market.
AI Agent Snapshot and Recommendation Intelligence answer two different questions, at two different levels.
AI agents are starting to interact with stores.
They search, retrieve, inspect products, check policies and access commerce capabilities. See what they actually do, not just that they visited.
AI agents are starting to interact with websites too.
They search, retrieve, inspect products, check policies and access commerce capabilities. This is early access for the website version, not the Shopify app.