跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
浏览器与自动化

shopify-admin-frequently-bought-together

Read-only: mines order history to find product pairs and triplets frequently purchased together, generating cross-sell and bundle recommendations.

40RTY-ai · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add 40RTY-ai/shopify-admin-skills --skill shopify-admin-frequently-bought-together

技能说明


name: shopify-admin-frequently-bought-together role: conversion-optimization description: "Read-only: mines order history to find product pairs and triplets frequently purchased together, generating cross-sell and bundle recommendations." toolkit: shopify-admin, shopify-admin-execution api_version: "2025-01" graphql_operations:

  • orders:query
  • products:query status: stable compatibility: Claude Code, Cursor, Codex, Gemini CLI

Purpose

Analyzes order history to discover which products are frequently purchased together. Calculates co-occurrence frequency, lift scores, and confidence metrics to generate data-driven cross-sell recommendations and bundle candidates. Read-only — no mutations.

Prerequisites

  • Authenticated Shopify CLI session: shopify store auth --store <domain> --scopes read_orders,read_products
  • API scopes: read_orders, read_products

Parameters

ParameterTypeRequiredDefaultDescription
storestringyes—Store domain
days_backintegerno180Order lookback window
min_supportintegerno3Minimum co-occurrence count to report a pair
max_resultsintegerno25Maximum product pairs to return
group_sizeintegerno2Pair size: 2 for pairs, 3 for triplets
collection_filterstringno—Limit to products in a specific collection
formatstringnohumanOutput format: human or json

Safety

ℹ️ Read-only skill — no mutations are executed. Safe to run at any time.

Workflow Steps

  1. OPERATION: orders — query Inputs: query: "created_at:>='<NOW - days_back days>'", first: 250, select lineItems { product { id, title } }, pagination cursor Expected output: All orders with product-level line items

  2. For each order with 2+ distinct products, generate all product pair combinations

  3. Build co-occurrence matrix:

    • Support = number of orders containing both products
    • Confidence(A→B) = P(B|A) = support(A,B) / support(A)
    • Lift = confidence(A→B) / P(B) — lift > 1.0 means positive association
  4. OPERATION: products — query (enrichment) Inputs: Product IDs from top pairs for titles, images, prices Expected output: Product details for display

  5. Rank pairs by lift score (descending), filter by min_support

GraphQL Operations

# orders:query — validated against api_version 2025-01
query OrderLineItems($query: String!, $after: String) {
  orders(first: 250, after: $after, query: $query) {
    edges {
      node {
        id
        lineItems(first: 50) {
          edges {
            node {
              product { id title }
              quantity
            }
          }
        }
      }
    }
    pageInfo { hasNextPage endCursor }
  }
}
# products:query — validated against api_version 2025-01
query ProductDetails($ids: [ID!]!) {
  nodes(ids: $ids) {
    ... on Product {
      id
      title
      vendor
      productType
      priceRangeV2 {
        minVariantPrice { amount currencyCode }
        maxVariantPrice { amount currencyCode }
      }
      totalInventory
      status
    }
  }
}

Session Tracking

Claude MUST emit the following output at each stage. This is mandatory.

On start, emit:

╔══════════════════════════════════════════════╗
║  SKILL: Frequently Bought Together           ║
║  Store: <store domain>                       ║
║  Started: <YYYY-MM-DD HH:MM UTC>             ║
╚══════════════════════════════════════════════╝

After each step, emit:

[N/TOTAL] <QUERY|MUTATION>  <OperationName>
          → Params: <brief summary of key inputs>
          → Result: <count or outcome>

On completion, emit:

For format: human (default):

══════════════════════════════════════════════
FREQUENTLY BOUGHT TOGETHER  (<days_back> days, <n> orders analyzed)
  Unique product pairs found: <n>
  Pairs meeting min_support:  <n>

  TOP PAIRS BY LIFT:
  #1  "<product A>" + "<product B>"
      Support: <n> orders  Lift: <n>x  Confidence: <pct>%

  #2  "<product A>" + "<product B>"
      Support: <n> orders  Lift: <n>x  Confidence: <pct>%

  BUNDLE CANDIDATES (high support + high lift):
    "<product A>" + "<product B>"  →  Suggested bundle price: $<n>

  Output: fbt_pairs_<date>.csv
══════════════════════════════════════════════

Output Format

CSV file fbt_pairs_<YYYY-MM-DD>.csv with columns: product_a_id, product_a_title, product_b_id, product_b_title, support, confidence_a_to_b, confidence_b_to_a, lift, combined_avg_price

Error Handling

ErrorCauseRecovery
THROTTLEDAPI rate limit exceededWait 2 seconds, retry up to 3 times
Single-item orders onlyStore with no multi-item ordersReport empty — suggest longer lookback window
Too many productsCombinatorial explosionLimit to top 500 products by order count

Best Practices

  • Use days_back: 180 or 365 for sufficient sample size.
  • Pairs with lift > 2.0 are strong bundle candidates.
  • Use results to create manual product bundles or configure upsell apps.
  • Cross-reference with top-product-performance to ensure paired items are high-performing.
  • Products with high confidence A→B but low confidence B→A suggest directional upsells (show B when A is in cart).