HasData Shopify

Shopify MCP Server

A hosted Shopify MCP server that gives Claude, Cursor, Windsurf and any other MCP client two read-only Shopify tools. Pull the product catalogue of any public Shopify storefront by its URL, with variants, SKUs and prices, and list the collections that organise it, all as structured JSON, with no app to install and no merchant token.

It reads the catalogue a signed-out visitor can see, on any classic Shopify storefront, whether it sits on a myshopify.com address or a custom domain. A headless store answers on its myshopify.com domain.

1,000 free credits every month, no card required, which is 100 Shopify calls at the 10-credit rate.

https://mcp.hasdata.com/mcp?apis=shopify

Contents

• What you need
• Quick start
• Example prompts
• Tools
• Errors and failure paths
• Pricing, free tier and limits
• Tool selection
• How it compares
• FAQ
• HasData links
• Development
• Contributing
• License

What you need

An MCP client and a HasData API key from the dashboard, free to create with no card, and the free tier covers about 100 calls a month at the 10-credit rate. This is a remote server, so the simplest path is a URL and an x-api-key header, with no container to run. A client that only speaks stdio reaches it through a thin launcher, published as @hasdata/shopify-mcp on npm and hasdata-shopify-mcp on PyPI, shown below.

Quick start

The server URL is the same for every client. We run it hands-on in Claude Code and Claude Desktop. The other blocks follow each client's own documented format for a remote server.

Clients with OAuth support can add the same URL as a connector and sign in without putting a key in a config file.

Claude Code

claude mcp add --transport http shopify "https://mcp.hasdata.com/mcp?apis=shopify" \

--header "x-api-key: HASDATA_API_KEY"

Claude Desktop

Settings, then Connectors, then Add custom connector, then paste https://mcp.hasdata.com/mcp?apis=shopify and sign in.

For the config-file route, Claude Desktop loads only local (stdio) servers, so it reaches a remote server through a stdio launcher. The @hasdata/shopify-mcp package is that launcher, and it reads the key from the environment. Add this to claude_desktop_config.json:

{

"mcpServers": { "shopify": { "command": "npx", "args": ["-y", "@hasdata/shopify-mcp"], "env": { "HASDATA_API_KEY": "YOUR_KEY" } } } }

For Python instead of Node, swap the launcher for the PyPI package, which uvx runs without a manual install:

{

"mcpServers": { "shopify": { "command": "uvx", "args": ["hasdata-shopify-mcp"], "env": { "HASDATA_API_KEY": "YOUR_KEY" } } } }

Cursor

~/.cursor/mcp.json for every project, or .cursor/mcp.json for one:

{

"mcpServers": { "shopify": { "url": "https://mcp.hasdata.com/mcp?apis=shopify", "headers": { "x-api-key": "HASDATA_API_KEY" } } } }

Windsurf

~/.codeium/windsurf/mcp_config.json. Windsurf calls the field serverUrl, not url:

{

"mcpServers": { "shopify": { "serverUrl": "https://mcp.hasdata.com/mcp?apis=shopify", "headers": { "x-api-key": "HASDATA_API_KEY" } } } }

VS Code

.vscode/mcp.json in the workspace:

{

"servers": { "shopify": { "type": "http", "url": "https://mcp.hasdata.com/mcp?apis=shopify", "headers": { "x-api-key": "HASDATA_API_KEY" } } } }

Example prompts

Each of these lands on one tool, or on two in sequence when the second needs a handle the first returns.

• List the collections on allbirds.com and tell me which ones hold the most products.
• Pull the first 250 products from this store and group them by product_type.
• Which variants in this store are out of stock right now?
• Find every product on this store that is discounted, comparing price against compare_at_price.
• Page through the shoes collection on this storefront and give me the price range per size.
• Compare the sock prices on these two Shopify stores.

A prompt that names a category rather than a handle takes two calls, one to list the collections and one to pull the products in the matching handle. The collection tool returns handle, and that value goes straight into the collection argument of the product tool.

Tools

Two tools, 10 credits per successful call. Both take a storefront URL and page through the results with limit and page, where limit accepts up to 250.

Get Shopify store products

hasdata_shopify_products_getProducts

A page of products from one storefront.

Returns a products array. Each product carries id, title, handle, body_html, vendor, product_type, tags, published_at, created_at, updated_at, an options array naming the axes the variants vary on, an images array, and a variants array.

The price lives on the variant, never on the product. A variant carries id, title, sku, price, compare_at_price, available, grams, position, requires_shipping, taxable, the option1 through option3 values and its own timestamps.

{

"id": 6889962537040, "title": "Anytime Ankle Sock - Basin Blue", "handle": "anytime-ankle-sock-basin-blue", "vendor": "Allbirds", "product_type": "Socks", "updated_at": "2026-09-09T05:25:09-07:00", "options": [{ "name": "Size", "position": 1, "values": ["S (W5-7)", "M (W8-10 / M8)", "L (W11 / M9-12)", "XL (M13-14)"] }], "variants": [ { "id": 40356485202000, "title": "S (W5-7)", "sku": "A10842U001", "price": "16.00", "compare_at_price": null, "available": false, "grams": 59, "position": 1 } ], "images": [ { "id": 36355227844688, "position": 1, "src": "https://cdn.shopify.com/s/files/1/1104/4168/files/A10842_S24Q1_Anytime_Ankle_Sock_Basin_Blue_A-1400x1400.png?v=1776183348", "width": 1400, "height": 1400 } ] }

Get Shopify store collections

hasdata_shopify_collections_getCollections

The collections that organise one storefront.

Returns a collections array. Each entry carries id, title, handle, description, image, products_count, published_at and updated_at.

This is the merchandising taxonomy as the store publishes it, which makes it the cheapest way to see how a competitor groups a catalogue before pulling any products. The handle is the join key into the product tool.

{

"id": 135995326544, "title": "Accessories", "handle": "womens-accessories", "description": "You know what they say: It's all in the details. Customize your look with planet-friendly face masks, hats, and more. ", "published_at": "2019-08-05T14:01:17-07:00", "updated_at": "2026-07-08T13:39:17-07:00", "image": null, "products_count": 28 }

Errors and failure paths

Plan for these rather than assuming a happy path.

`price` and `compare_at_price` are strings, not numbers. They arrive as "16.00", exactly as the storefront publishes them. Convert before you compare or sum, because string ordering puts "9.00" above "16.00".

A product has no price of its own. Anything about cost has to go through the variants array, and a product with a size or colour axis usually has several prices. Reading the first variant and calling it the price is the most common mistake here.

`compare_at_price` is null when nothing is discounted, so a discount check is a null test first and a comparison second.

`available` is per variant and reflects the moment of the call. A product is not out of stock, a variant is, and stock moves. Two calls minutes apart can disagree, which is the point when you are monitoring, and a trap when you are diffing catalogues.

A collection can report `products_count` of zero. Stores leave empty, staged and seasonal collections published, so an empty collection is normal rather than a failed call.

Stores publish things that are not for sale. Internal, retired and staging items sit in the public catalogue on plenty of stores, sometimes flagged in the title and sometimes not. Filter on what you need instead of trusting that every row is a live product.

`tags` are whatever the merchant wrote. Some stores use them as plain keywords, others push namespaced metafield strings into them. Treat the array as free text.

`body_html` is HTML. Strip it before you index or embed the description.

`collection.image` is often null, and so is featured_image on a variant. Fall back to the product images array.

A URL that is not a classic Shopify storefront still answers 200 and still bills. There is no products array in that response and an error string in its place, while requestMetadata.status stays ok. Test for the array before you read it, because the shape changes rather than the status.

**A