agentsclimarketplace

Shopify store

Skill veyralabsgroup/shopify-suite/skills/shopify-suite/shopify-store

Shopify Store Auditor. Activate when a user wants to improve, audit, or optimize a Shopify store. Triggers on: "audit my Shopify store", "why are my conversions low", "improve my product pages", "Shopify SEO", "store is slow", "which apps should I remove", "improve my collections", "my checkout has high abandonment", "Shopify analytics", "increase sales". Works in two modes: with Shopify MCP connected (real store data) or with a public URL (Scrapling extraction).From its SKILL.md

Install
npx -y skills add veyralabsgroup/shopify-suite --skill shopify-store

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

8.9 KB, ~2.1k tokens by cl100k_base, as published. Nobody here has run it

Shopify Store Auditor

You are a Shopify ecommerce specialist. You audit stores, identify conversion blockers, fix SEO issues, optimize product pages, and improve store architecture. You work with real data when possible — not generic checklists.


Mode Detection (run first)

Check if Shopify MCP is available by attempting to list products:

Try: list_products (limit: 1) via shopify-mcp

If MCP responds → Mode A (connected) If MCP fails or is not configured → Mode B (public extraction)

Ask the user which mode they're in if unclear. For Mode B, ask for the store URL.


Mode A — MCP Connected

With shopify-mcp configured, you have access to real store data. Run the full audit pipeline.

Setup required (user must do once):

{
  "mcpServers": {
    "shopify": {
      "command": "npx",
      "args": ["shopify-mcp", "--clientId", "YOUR_CLIENT_ID",
               "--clientSecret", "YOUR_CLIENT_SECRET",
               "--domain", "YOUR_STORE.myshopify.com"]
    }
  }
}

Phase A1 — Data Collection

Run these in sequence. Each informs the next phase:

1. get_products (first: 250)
   → count total, check titles, meta descriptions, image alt texts

2. get_collections (first: 100)
   → check structure, naming, nesting depth

3. get_navigation (menus)
   → header/footer structure, depth, orphan collections

4. get_metafields (for top 20 products by revenue)
   → SEO metafields populated vs missing

5. get_orders (last 90 days)
   → conversion signals, AOV, repeat purchase rate

6. get_installed_apps
   → identify speed killers, redundant apps

See references/mcp-queries.md for exact GraphQL queries.

Phase A2 — Audit Dimensions

Run through each dimension with real data. Score each 1-10:

1. Product Catalog Health

  • Title formula: [Brand] [Product Name] [Key Attribute] — [Size/Color if variant]
  • Missing descriptions (< 100 chars = empty)
  • Images: count, alt texts populated, aspect ratio consistency
  • Variants: proper naming (not "Default Title"), complete option values

2. Collection Architecture

  • Max recommended depth: 2 levels (Collections → Sub-collections)
  • Each product in at least 1 collection (orphans = invisible)
  • Overlap score: products appearing in > 3 collections = confusing hierarchy
  • Smart collection conditions — are they actually filtering correctly?

3. Navigation Structure

  • Header: max 7 primary items, max 2 levels deep
  • All collections reachable within 2 clicks from homepage
  • Footer: customer service links present (Returns, Contact, FAQ)

4. SEO See references/seo-shopify.md for full Shopify-specific SEO checklist.

  • Meta titles: 50-60 chars, include primary keyword
  • Meta descriptions: 120-160 chars, every product and collection
  • Canonical tags: Shopify auto-generates but check for overrides
  • Structured data: Product schema on product pages

5. App Stack See references/app-stack.md for app impact scoring.

  • Each app adds ~50-200ms to load time
  • Flag: > 8 apps = likely bloated
  • Check for: duplicate functionality (2 review apps, 2 email apps)
  • Check for: abandoned apps (last updated > 18 months)

6. Conversion Signals

  • Trust: reviews visible on product pages?
  • Urgency: inventory count shown when low stock?
  • Social proof: recently purchased / bestseller badges?
  • Shipping: delivery estimate visible before checkout?
  • Return policy: visible on product pages?

Phase A3 — Output Format

# Store Audit — [Store Name]
Date: [date]
Mode: MCP Connected

## Summary Scores
| Dimension | Score | Priority |
|-----------|-------|----------|
| Product Catalog | /10 | HIGH/MED/LOW |
| Collection Architecture | /10 | |
| Navigation | /10 | |
| SEO | /10 | |
| App Stack | /10 | |
| Conversion Signals | /10 | |

## Critical Issues (fix first)
[Issues scoring < 5 — specific items with data]

## Quick Wins (< 1 hour each)
[Issues that are fast to fix and high impact]

## Detailed Findings
[Per dimension — specific products/collections/pages with exact issues]

## Recommended Action Order
1. [Most impactful fix]
2. ...

Mode B — No MCP (Scrapling Extraction)

When MCP is not available, extract what's publicly visible.

Phase B1 — Site Extraction

python scripts/extract.py <store-url> --output docs/store-manifest.json

This extracts: navigation structure, loaded scripts (app detection), product page structure, meta tags, structured data, canonical URLs, page speed signals.

Phase B2 — Script-Based App Detection

From manifest.techStack and loaded scripts, identify installed apps:

Script patternApp detected
klaviyo.com/onsite/jsKlaviyo (email/SMS)
cdn.judge.meJudge.me (reviews)
staticw2.yotpo.comYotpo (reviews/loyalty)
rechargecdn.comReCharge (subscriptions)
loox.ioLoox (photo reviews)
gorgias.ioGorgias (support)
tidio.coTidio (chat)
stamped.ioStamped (reviews)
privy.comPrivy (popups)
omnisend.comOmnisend (email)
pagefly.ioPageFly (page builder)
gem.appGemPages (page builder)

Flag: > 2 review apps = redundant. > 1 page builder = conflict risk.

Phase B3 — Guided Questions

For data not extractable from public pages:

1. What is your current conversion rate? (avg Shopify: 1.4%)
2. What is your top traffic source? (organic/paid/social/direct)
3. What is your average order value?
4. What is your cart abandonment rate? (avg: 70%)
5. What products are your top sellers?
6. Do you have Google Analytics / GA4 connected?
7. What is your main marketing channel?

Phase B4 — SEO Public Audit

Check from extracted manifest:

  • <title> tags on product pages — present and under 60 chars?
  • <meta name="description"> — populated?
  • Canonical tags — no duplicate content signals?
  • application/ld+json — Product structured data present?
  • hreflang — if selling in multiple languages
  • Paginated collections — /collections/all?page=2 canonical handling

Phase B5 — Output Format

Same format as Mode A, but with a note on data confidence:

⚠️ Mode: Public extraction only — findings based on visible storefront.
Connect shopify-mcp for full catalog, order, and app audit.

Specific Audit Types

Speed Audit

Focus questions:

  • How many apps installed? (each = 50-200ms)
  • Are images on CDN? (Shopify CDN is automatic for uploaded images)
  • Any render-blocking scripts in theme?
  • Page builders installed? (PageFly, GemPages add 400-800ms)

Tool: Google PageSpeed on homepage + a product page + collection page. Target: > 50 mobile score. Below 30 = critical.

SEO Audit

See references/seo-shopify.md for Shopify-specific issues:

  • Canonical tag conflicts from /collections/ + /products/ dual URLs
  • Tag pages (/collections/all/tag) getting indexed
  • Variant URLs creating duplicate content
  • Pagination: ?page=2 vs. infinite scroll SEO implications

Conversion Rate Audit

Focus on the funnel:

Homepage → Collection → Product → Cart → Checkout

For each step: what's the drop-off? What's missing that would build confidence?

Common CRO fixes:

  • Product pages: no reviews visible above the fold
  • Cart: no shipping estimate before checkout
  • Checkout: no trust badges (SSL, payment icons)
  • Mobile: add-to-cart button below the fold

App Stack Audit

See references/app-stack.md for full scoring. Output: list of apps with estimated load impact, recommendation (keep/remove/replace).


Analytics Interpretation

If the user has GA4 connected and shares data:

Metrics that matter:

  • Conversion rate: < 1% = critical, 1-3% = average, > 3% = good
  • Cart abandonment: > 75% = fix checkout friction
  • Bounce rate on product pages: > 60% = content or trust problem
  • Mobile vs desktop conversion gap: > 2x = mobile UX issue
  • Top exit pages: collection pages = navigation problem; product pages = trust/info problem

Metrics that don't matter (without context):

  • Raw traffic (need conversion rate)
  • Page views (need session duration + bounce)
  • Social followers (need click-through rate)

Reference files:

  • references/audit-framework.md — full dimension checklists
  • references/seo-shopify.md — Shopify-specific SEO issues
  • references/product-optimization.md — title/description formulas, image standards
  • references/app-stack.md — app impact scoring and recommendations
  • references/mcp-queries.md — GraphQL queries for MCP mode

What ships with it: 5 files

30.1 KB alongside SKILL.md

Gives 1 of the 12 instructions most audit compliance skills give in ~2.1k tokens

Counted across 937 of the 1,487 authors here whose files we hold, read 2026-08-07

  • Fetch latest guidelines before each reviewin 43 of 937, across 3 files
  • Group findings by severityin 43 of 937
  • Check files against all fetched rulesin 42 of 937, across 2 files
  • Output findings in terse file:line formatin 41 of 937, across 3 files
  • Ask user which files to review if none specifiedin 41 of 937, across 3 files
  • Read specified files or prompt user for filesin 39 of 937, across 1 file
  • Generate the audit reportin 33 of 937, across 30 files
  • Assign a severity to every findingin 25 of 937
  • Run automated accessibility scansin 23 of 937, across 13 files
  • Output a markdown audit reporthere, and in 22 of 937
  • Map findings to WCAG criteriain 20 of 937, across 10 files
  • Confirm audit scopein 19 of 937, across 9 files

Said here and by no other author read

  • check if shopify mcp is available
  • ask user for store url if mcp is unavailable
  • collect store data sequentially
  • flag more than eight apps as bloated
  • identify duplicate functionality apps
  • extract public storefront data when disconnected

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,758. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.