Shopify Storefront MCP Server
AI shopping agent for Shopify stores: product search, cart management, and policy queries directly via the MCP standard.
- Skill Road
- Shopify Storefront MCP Server
Categories
Description
The Shopify Storefront MCP Server is Shopify's official Model Context Protocol server for store-level commerce data. It connects AI assistants and shopping agents to the product catalog, cart, and policies of a specific Shopify store. Unlike the Shopify Dev MCP Server — which accesses developer documentation and API schemas — the Storefront MCP Server is designed for end-customer scenarios: a buyer asks a question about products, the agent searches the selected store's catalog, adds items to the cart, and guides the shopper through checkout. The official documentation page is https://shopify.dev/docs/apps/build/storefront-mcp.
What is the Shopify Storefront MCP Server?
The Storefront MCP Server provides, according to the provider, structured access to Shopify's commerce data: product search, cart operations, and queries about store policies. Each Shopify store has its own MCP endpoint at https://{shop}.myshopify.com/api/mcp. This endpoint is store-specific and covers all storefront commerce capabilities. The Shopify app that acts as the MCP client routes the AI model's requests to this endpoint; the Chat UI is embedded into the store's theme as a Shopify theme extension.
Architecture: MCP client, MCP server, and Chat UI
According to the provider, the system uses a client-server architecture following the Model Context Protocol: the MCP client is the Shopify app that connects with AI models and passes their requests to MCP servers. The MCP server is an API endpoint that provides structured access to commerce data such as products, cart operations, and customer information. The Chat UI is a Shopify theme extension that presents a customer-facing chat window. The backend acts as the MCP client while the frontend delivers the customer experience. This approach enables any AI model to be connected to Shopify without custom integration work.
Available tools
The Shopify Storefront MCP Server provides the following tools according to the official documentation:
search_catalog: Searches the store's product catalog for items matching the customer's needs. Supports free-text search, buyer signals (country, language, currency, purchase intent), filters by category and price range, and cursor-based pagination. This tool conforms to the UCP catalog specification and is available at https://{shop}.myshopify.com/api/ucp/mcp.
lookup_catalog: Retrieves products or variants by identifier. Useful when product or variant IDs are already available from search results, saved lists, or deep links. Supports up to ten IDs per request, buyer localization signals, and downstream filters.
get_product: Retrieves full details for a single product, with optional interactive variant selection. Appropriate when a buyer has selected a product and needs detailed information for a purchase decision. Supports variant selection and preference relaxation order.
search_shop_policies_and_faqs: Answers questions about the store's policies, products, and services to build customer trust. According to Shopify, only the provided answer should be used to form the response; external information may be inaccurate.
get_cart: Retrieves the current contents of a cart, including item details and checkout URL. Requires a cart ID.
update_cart: Updates quantities of items in an existing cart or adds new items. Creates a new cart if no cart ID is provided. Setting quantity to 0 removes an item.
Distinction: Storefront MCP vs. Dev MCP vs. Customer Accounts MCP
Shopify operates multiple MCP servers for different use cases. The Storefront MCP Server serves end-customer shopping scenarios: product search, cart, policies. It requires no authentication for the storefront endpoint. The Dev MCP Server (@shopify/dev-mcp) targets exclusively developers: it provides access to Shopify developer documentation and API schemas, not store data. The Customer Accounts MCP Server (https://{shopDomain}/customer/api/mcp) enables authenticated customer-specific actions such as order tracking and returns; it requires OAuth 2.0 with PKCE and access to protected customer data approved via the Shopify Partner Dashboard.
Setup for developers
Integrating a Storefront MCP Server into a custom Shopify app requires the following components: a Shopify app (Node.js recommended), a Shopify Partner account, a development store, and basic knowledge of MCP and Shopify app development. The endpoint is store-specific and takes the form https://{shop}.myshopify.com/api/mcp. For UCP catalog tools, the separate endpoint https://{shop}.myshopify.com/api/ucp/mcp applies, along with an agent profile header. The official guide recommends starting with the Shopify AI Toolkit at shopify.dev/docs/apps/build/devmcp for the developer onboarding path.
Security notes
The Storefront MCP Server requires no authentication for public store data according to the official documentation — product catalog and policies are publicly accessible by design. For the Customer Accounts MCP Server, OAuth 2.0 (authorization code flow with PKCE) is required, and the developer must request Level 2 protection for customer data (PII) through the Shopify Partner Dashboard. Shopify notes that individual stores may restrict access; therefore the integration should always be tested with the specific development store. By using the Shopify MCP servers, the developer agrees to Shopify's API terms of use according to the provider. Data retrieved by the AI agent is embedded in model requests; the AI model used and its privacy policy should be reviewed accordingly.
FAQ
Do end customers need a Shopify account? No. The Storefront MCP Server provides public product data; a Shopify customer account is only required for the Customer Accounts MCP Server, which manages order history and personal data.
Is the Storefront MCP Server free? The MCP endpoints are available as part of Shopify's Storefront API without additional cost. Costs arise from Shopify Partner plans, the chosen AI model provider, and potentially from Shopify app fees. Current pricing information is available at shopify.dev.
Is there a public GitHub repository? According to research conducted on 2026-09-08, no public repository exists under github.com/Shopify for this specific MCP server. The server is a cloud-hosted, store-specific endpoint without a separate open-source package.
What is the difference from the Shopify Storefront API? The Storefront API is a GraphQL API for direct queries. The Storefront MCP Server builds on the MCP standard and enables AI assistants to access the same commerce data with structured, tool-based calls, without the AI side needing to formulate GraphQL queries directly.
Requirements
Shopify Partner account, Shopify app (Node.js recommended), development store, MCP-compatible AI client or custom app integration.
Installation instructions
Quick start
Configure the store-specific MCP endpoint https://{shop}.myshopify.com/api/mcp in your Shopify app as the MCP client. For UCP catalog tools, use the endpoint https://{shop}.myshopify.com/api/ucp/mcp. Full guide: shopify.dev/docs/apps/build/storefront-mcp.
What the Shopify Storefront MCP Server is
According to the provider, the Shopify Storefront MCP Server is an officially maintained interface built on the Model Context Protocol, an open standard that defines how applications hand structured context and tools to AI models. Instead of hand-coding every AI integration, an AI-enabled application connects over MCP to one or more servers, each exposing a clearly scoped set of capabilities. In Shopify's case, the Storefront MCP server gives an AI assistant access to a store's catalog, cart, and policies, letting shoppers search, ask questions, and buy using natural language. A companion Customer Accounts MCP server covers order tracking, returns, and account information. Together with an MCP client built into the Shopify app and a chat-facing theme extension, these two servers form the backbone of an AI shopping experience that lives directly inside a merchant's storefront.
Step 1: Prerequisites for using it
To use the Storefront MCP server in production, a merchant needs an active Shopify store and a basic understanding of Shopify app development, since the server is wired up through a Shopify app that acts as the MCP client. That app connects to the server and relays its responses into a chat interface delivered as a theme extension inside the store's design. Developers should review Shopify's official documentation before building, because permission scopes, data models, and supported actions can evolve as the feature matures. An AI model or compatible agent capable of issuing MCP requests is also required, since the server itself supplies structured data and actions rather than a language model.
Step 2: Setup and architecture
The setup follows the standard MCP client-server pattern. The Shopify app serves as the built-in MCP client and connects to the Storefront MCP server and, optionally, to the Customer Accounts MCP server. On the server side, structured endpoints handle product discovery, cart management, and checkout flows. The chat interface is delivered as a theme extension embedded in the storefront, giving customers a chat window where they can ask about products in natural language, add items to their cart, or check order status. Conversations stay coherent across multiple messages, and, according to the provider, responses stream in to keep the experience feeling natural. The chat widget's appearance can be styled to match the store's branding.
Security and practical value
Because the server touches sensitive commerce data such as product details, cart contents, and potentially customer information, access scopes should be granted deliberately and kept minimal. It also matters that only apps authorized by the specific store can reach the MCP servers, and that order and account data flows exclusively through the more tightly controlled Customer Accounts endpoints rather than the general storefront path. In practice, stores with large catalogs or complex return policies benefit most, since an AI chat can automate routine questions and take pressure off support teams. At the same time, the approach is tied to Shopify's own ecosystem: merchants running storefronts outside Shopify cannot use this specific pair of servers directly and would need a more general MCP implementation or a custom integration instead. The quality of the AI's answers also depends heavily on the connected language model and its configuration, so ongoing testing and tuning remain necessary before treating such a setup as fully production ready.
Authentication
No authentication for public storefront data (products, policies, cart). OAuth 2.0 with PKCE for the Customer Accounts MCP Server (orders, customer data).
Required access permissions
Read access to public catalog and policy data; write access to carts. The Customer Accounts MCP Server requires Level 2 protected customer data (PII) approval.
Transmitted or stored data
The server is a cloud-hosted, store-specific endpoint operated by Shopify. Retrieved commerce data is embedded in the AI client's model requests. Customer data in the Customer Accounts MCP is protected by Shopify's privacy policies and OAuth.
Security risks
Storefront endpoint: low risk profile for public data. Customer Accounts MCP: PII access requires OAuth 2.0 and Shopify approval. Retrieved data flows into model requests; review the AI model used and its privacy policy.
License and costs
- License
- Not recorded yet.
- Cost
- free
MCP endpoints available at no extra cost as part of the Storefront API. Costs arise from Shopify Partner plan, chosen AI model provider, and potentially app fees. Current information at shopify.dev.
Alternatives
Not recorded yet.
At a glance
- Provider
- Shopify
- Status
- Official server
- Deployment
- Remote
- Current version
- Not recorded yet.
- Last reviewed
- 08.09.2026
Repository and documentation
Categories
Supported clients
Not recorded yet.
Related guides
Guides and background related to this entry.
No related guides yet.