Cart Now Has a Public API, SDKs, and an MCP Server
We just shipped something we've been working toward for a while: Cart's entire e-commerce intelligence platform is now accessible through a public API, open-source SDKs, and a Model Context Protocol (MCP) server.
If you work with e-commerce data - stores, products, ads, traffic, suppliers - you can now pull it into your code, your AI agent, or Claude directly.
The API
Cart has tracked thousands of Shopify stores for years. Until now, that data lived behind our dashboard. Not anymore.
The REST API gives you programmatic access to everything:
- Stores - look up any Shopify store, get traffic estimates, tech stack, and metadata
- Products - search across millions of products, filter by category, price, or store
- Ads - see what stores are running on Facebook and Google
- Traffic - monthly visitor estimates, traffic sources, and trends
- Suppliers - find who supplies products to specific stores
Full docs at docs.usecart.com.
Free tier included. 100 requests per day, no credit card required.
Two SDKs, Zero Dependencies
We shipped SDKs for both Python and TypeScript:
pip install usecart
npm install @usecart/sdk
Both were built with zero runtime dependencies. The Python SDK uses only stdlib urllib. The TypeScript SDK uses just
fetch. No bloat, no transitive dependency chains, no supply chain risk.
from usecart import Cart
cart = Cart("cart_sk_your_key")
store = cart.stores.get("gymshark.com")
print(store.monthly_visits)
import { Cart } from "@usecart/sdk";
const cart = new Cart("cart_sk_your_key");
const store = await cart.stores.get("gymshark.com");
console.log(store.monthlyVisits);
MCP Server: E-Commerce Data Inside Claude
This is the part that excites us most.
Claude can now look up any Shopify store, search products, analyze niches, and compare competitors - all through our
MCP server. No code needed.
Add one line to your Claude config:
npx @usecart/mcp-server
Then just ask questions in natural language:
- "What's Gymshark's estimated monthly traffic?"
- "Find Shopify stores selling yoga mats with over 100k monthly visits"
- "Who supplies products to this store?"
- "Compare these three competitors side by side"
Claude calls the Cart API behind the scenes and gives you a structured answer. It turns conversational questions into
e-commerce research.
Claude Code Skills
We also shipped a set of e-commerce skills for Claude Code - slash commands that pull live intelligence into your
terminal while you work:
- /store-lookup - instant store profile and metrics
- /niche-research - analyze a market segment with real data
- /competitor-report - side-by-side competitor comparison
- /find-suppliers - discover suppliers for any product category
These are purpose-built workflows, not generic prompts. Each one knows what data to fetch, how to structure the
analysis, and what matters for the specific use case.
Why We Built This
E-commerce research has been manual for too long. You visit a dozen tools, copy numbers into spreadsheets, and try to
piece together a picture. AI agents should be able to do this work, but they need access to the data.
That's what this release is about. Whether you're building a dropshipping tool, running competitive analysis, or just
curious about a store - the data is now one API call away.
Get started: <https://docs.usecart.com>
Free tier: 100 requests/day, no credit card
Open source: SDKs and MCP server are on npm and PyPI