Using MCP for commerce data
How to design Model Context Protocol tools for commerce so agents answer questions instead of stitching fragments.
The tempting mistake is to mirror REST endpoints as tools one-for-one. That produces a chatty server: the agent burns context stitching partial answers and gets a fresh opportunity to hallucinate the join at every step.
Design tools around questions instead. “Where is this cheapest, landed?” is a tool. “Get offers” is a fragment. A data server is worth more the more work it completes before responding.
Declare cost per tool. One prompt fans into dozens of calls with no human approving any of them, so a metadata lookup and a live cross-seller comparison must not bill the same or agents learn to abuse the cheap one.
Keep reading
What is commerce intelligence?
Commerce intelligence is the layer above price monitoring: an index of what exists, who sells it, at what, and what that means.
Why percentile beats percent-off
A percentile against a product's own history is a defensible measure of discount depth; percent-off a reference price is not.
Product matching and why it fails quietly
How cross-seller product matching works, the three ways it fails, and why published match confidence matters.
GTIN, UPC, EAN and ASIN
How product identifiers relate, why check digits matter, and what to do when there is no identifier at all.
Data provenance in commerce data
Why every price should carry its origin and a confidence weight, and what goes wrong when it does not.