OVERVIEW

Shopify powers over 4 million stores globally. It exposes two APIs: the Admin API (REST + GraphQL) for store management, and the Storefront API (GraphQL) for customer-facing experiences. Both are powerful foundations for AI augmentation.

This skill gives AI agents the knowledge to query, operate, and automate Shopify — across every role in the organisation — without hallucinating endpoints or field names.

What your agent can do

• Query products, orders, customers, inventory, collections, and analytics
• Generate AI-powered product descriptions, SEO copy, and alt text
• Surface revenue dashboards, trend analysis, and stock alerts for owners
• Audit SEO issues, auto-tag products, and reorder collections for merchandisers
• Instantly look up orders and draft customer service responses
• Segment customers and generate campaign copy for marketing
• Monitor unfulfilled orders, flag fraud, and route fulfillments for operations

THE 2NTH MODEL

The 2nth model is simple: every person who works with a Shopify store gets their own AI partner. The AI never makes decisions — it surfaces data, drafts content, and catches things humans miss. The human always has the final say.

ARCHITECTURE

The recommended integration uses a Cloudflare Worker as an MCP server — it proxies the Shopify Admin API, generates content via Workers AI, and exposes role-specific tools to the AI client.

Authentication

# Admin API (Private/Custom App)
X-Shopify-Access-Token: shpat_xxxxx

# Storefront API
X-Shopify-Storefront-Access-Token: xxxxx

# Base URLs
https://{store}.myshopify.com/admin/api/2024-10/{resource}.json   # REST
https://{store}.myshopify.com/admin/api/2024-10/graphql.json      # GraphQL

MCP Tool Registration (per role)

const ROLE_TOOLS = {
  owner:        ['get_revenue_dashboard', 'get_top_products', 'get_customer_growth', 'get_inventory_value'],
  merchandiser: ['list_products', 'update_product', 'manage_collection', 'get_seo_audit', 'auto_tag_products'],
  content:      ['get_product', 'update_product_description', 'generate_alt_text', 'update_seo_metadata'],
  support:      ['search_orders', 'get_order_status', 'search_customers', 'draft_response', 'create_return'],
  marketing:    ['get_sales_report', 'get_customer_segments', 'generate_campaign_copy', 'get_channel_performance'],
  operations:   ['list_unfulfilled_orders', 'get_inventory_levels', 'create_fulfillment', 'flag_fraud_risk'],
};

SHOPIFY APIS

GraphQL is preferred for complex queries, metafields, and bulk operations. REST is simpler for straightforward CRUD. Both are available on all plans.

Admin REST — Key Resources

GET  /admin/api/2024-10/products.json?limit=50&status=active
GET  /admin/api/2024-10/products/{id}.json
POST /admin/api/2024-10/products.json
PUT  /admin/api/2024-10/products/{id}.json

GET  /admin/api/2024-10/orders.json?status=open&limit=50
GET  /admin/api/2024-10/orders/{id}.json

GET  /admin/api/2024-10/customers/search.json?query=email:[email protected]
GET  /admin/api/2024-10/inventory_levels.json?location_ids=1234

Admin GraphQL — Products

{
  products(first: 20, query: "status:active") {
    edges {
      node {
        id title handle description tags status
        variants(first: 10) {
          edges { node { id price sku inventoryQuantity } }
        }
        images(first: 5) {
          edges { node { url altText } }
        }
        seo { title description }
      }
    }
    pageInfo { hasNextPage endCursor }
  }
}

Admin GraphQL — Orders

{
  orders(first: 30, query: "fulfillment_status:unfulfilled") {
    edges {
      node {
        id name createdAt
        totalPriceSet { shopMoney { amount currencyCode } }
        customer { firstName lastName email }
        lineItems(first: 10) {
          edges { node { title quantity variant { sku inventoryQuantity } } }
        }
        shippingAddress { address1 city province zip country }
        fulfillments { trackingInfo { number url company } status }
      }
    }
  }
}

Rate Limits

Webhooks

OWNER AI

The store owner sets strategy, makes pricing decisions, and chooses brand direction. Their AI partner handles daily data retrieval so they can focus on what only they can do.

MCP Tools

Example Conversation

MERCHANDISER AI

The merchandiser curates collections, decides what to feature, and approves product positioning. Their AI partner handles the analytical legwork — surfacing what needs attention before it becomes a problem.

MCP Tools

SEO Audit Query

{
  products(first: 250) {
    edges {
      node {
        id title
        seo { title description }
        images(first: 1) { edges { node { altText } } }
      }
    }
  }
}

Example Conversation

CONTENT AI

The content creator owns the brand voice, makes creative decisions, and approves all copy. Their AI partner handles first drafts — so they spend time on craft, not data entry.

MCP Tools

Content System Prompt Pattern

You are a copywriter for NoHa, a luxury South African furniture brand.
Brand voice: sophisticated, understated, nature-inspired. Never salesy.
Lead with the emotional benefit. Include materials and dimensions.
Use sensory language. Keep under 150 words for product descriptions.

Workers AI — Content Generation

const result = await env.AI.run('@cf/meta/llama-3.1-8b-instruct', {
  messages: [
    { role: 'system', content: 'You are a luxury e-commerce copywriter. Be concise and compelling.' },
    { role: 'user', content: `Write a product description for "${product.title}".
      Brand: luxury South African furniture. Voice: sophisticated, nature-inspired.
      Include: ${product.tags.join(', ')}.
      Keep under 150 words. Lead with emotion, end with specs.` }
  ]
});

Example Conversation

SUPPORT AI

Customer service handles escalations, approves refunds, and makes relationship calls. Their AI partner handles the lookup and drafting work — so they respond faster without losing the human touch.

MCP Tools

Order Lookup Query

{
  orders(first: 1, query: "name:#1042") {
    edges {
      node {
        id name financialStatus fulfillmentStatus
        totalPriceSet { shopMoney { amount } }
        customer { firstName lastName email }
        lineItems(first: 10) { edges { node { title quantity } } }
        fulfillments { trackingInfo { number url company } status }
      }
    }
  }
}

Example Conversation

MARKETING AI

The marketing manager decides campaign strategy, approves messaging, and allocates budget. Their AI partner segments the audience, generates copy variants, and surfaces the data that makes campaigns perform.

MCP Tools

Customer Segmentation Query

{
  customers(first: 50, query: "orders_count:>3", sortKey: TOTAL_SPENT, reverse: true) {
    edges {
      node {
        id firstName lastName email
        ordersCount totalSpent tags
        lastOrder { id name createdAt }
      }
    }
  }
}

Example Conversation

OPERATIONS AI

Operations handles exceptions, selects carriers, and manages the warehouse team. Their AI partner monitors the queue, flags issues before they escalate, and pre-builds fulfillment requests for human approval.

MCP Tools

Unfulfilled Orders Query

{
  orders(first: 30, query: "fulfillment_status:unfulfilled") {
    edges {
      node {
        id name createdAt
        totalPriceSet { shopMoney { amount currencyCode } }
        customer { firstName lastName email }
        lineItems(first: 10) {
          edges { node { title quantity sku variant { inventoryQuantity } } }
        }
        shippingAddress { address1 city province country }
      }
    }
  }
}

Example Conversation

INSTALL

Install this skill into your AI agent with a single command. The skill gives your agent complete Shopify API knowledge, all role patterns, query references, and content generation prompts.

What's Included

• SKILL.md — Full Admin API reference, role patterns, rate limits, webhooks, gotchas
• references/queries.md — Complete query library: products, orders, customers, inventory, analytics, bulk ops
• references/roles.md — Detailed role patterns for all 6 roles with MCP tools and example conversations
• references/ai-integration.md — Cloudflare Worker proxy, Workers AI content generation, webhook handlers, wrangler config

Wrangler Config

name = "shopify-ai"
compatibility_date = "2026-03-01"
compatibility_flags = ["nodejs_compat"]

[ai]
binding = "AI"

[vars]
SHOPIFY_STORE = "your-store-name"

# Secrets (run in your terminal):
# npx wrangler secret put SHOPIFY_ACCESS_TOKEN