Claude Prompts for Shopify: A Library of 40+ Prompts

Key takeaways

• This is a copy-paste library of 40+ Claude prompts for Shopify operators and developers, grouped by job: theme/Liquid dev, PDP copy, SEO and metafields, CRO, analytics, support, debugging, and migration.
• Claude returns better Shopify output when you give it context, constraints, and the exact format you want back - paste your real theme code, your brand voice, and your acceptance criteria.
• For Liquid and theme work, tell Claude the Shopify-specific rules it must follow (Online Store 2.0 sections, schema settings, no liquid in JSON) or it will write generic templating.
• Claude writes code, but it does not touch your store. To turn a prompt into a live theme change, pair it with a tool like Fudge that writes the code into your theme and previews it before publish.

This is a working prompt library for Shopify operators and developers who use Claude. Every prompt below is written to drop straight into a Claude chat. Replace the bracketed placeholders with your real values, paste in any code or copy Claude needs, and send.

The prompts are grouped by the job you are doing. Skip to the section you need.

Why you can trust us

We have been in the Shopify space for over four years and built Fudge, an AI page builder and storefront editor used by hundreds of Shopify merchants. Jacques, our co-founder, has more than 15 years of development experience and has shipped storefront code for dozens of brands. We use Claude daily for theme work, copy, and debugging, and the prompts here are the patterns that actually hold up on real stores.

For a wider view of how these tools fit together, see our Shopify AI toolkit and our take on AI-first Shopify development.

How to get good Shopify answers from Claude

Three habits separate a useful answer from a generic one.

Give Claude the real artifact. Paste the actual Liquid section, the real product data, the exact error. Claude reasons over what you give it. A vague description produces vague code.

State the Shopify constraints. Claude knows general web development. It needs reminding that themes use Online Store 2.0 sections, that settings live in a {% schema %} block, that you cannot put Liquid inside a JSON template, and that metafields have specific types. Spell these out.

Demand a format. Ask for a single fenced code block, a table, or a numbered checklist. Without a format instruction Claude pads answers with prose you have to trim.

Every prompt below already bakes these in. Now the library.

Theme and Liquid development prompts

These assume you are working in a Dawn-style Online Store 2.0 theme. Paste the relevant section or snippet alongside each prompt.

Build a new section from scratch.

You are a senior Shopify theme developer working in an Online Store 2.0
Dawn-based theme. Build a new section called "[section name]" that does
the following: [describe the layout and behavior].

Requirements:
- Use a {% schema %} block with settings so a merchant can edit
  [list the editable fields] from the theme editor.
- Make every text and image field a schema setting, not hard-coded.
- Add a "presets" entry so it can be added from the section picker.
- Use semantic HTML and BEM-style class names prefixed with the section name.
- Do not use inline styles; output a matching <style> block scoped to the section.

Return one complete .liquid file in a single code block.

Tip: name the section in the prompt exactly as you want the file named so you can save the output directly.

Explain unfamiliar Liquid.

Explain what this Shopify Liquid does, line by line, in plain English.
Flag anything that could break on an empty cart, a product with no
variants, or a collection with zero products.

[paste the Liquid snippet]

Tip: this is the fastest way to safely inherit a theme you did not write.

Refactor a section to add a schema setting.

Here is a Shopify section. Add a theme-editor setting called
"[setting label]" of type [text/range/color/select] that controls
[what it controls]. Wire it into the existing markup, keep all current
behavior intact, and return the full updated section.

[paste the section]

Convert hard-coded content into block settings.

This section hard-codes its [testimonials / logos / feature cards].
Convert them into repeatable blocks so a merchant can add, remove, and
reorder them in the theme editor. Define the block type and its settings
in the schema, then loop over section.blocks in the markup. Return the
full section.

Write a reusable snippet.

Write a Shopify snippet called [snippet name] that renders [what it
renders]. It should accept these parameters via the render tag:
[list parameters and their types]. Include a comment block at the top
documenting the parameters and an example {% render %} call. Return the
snippet file only.

Add a metafield-driven element to a product template.

On the product page, I want to display a [size chart / care guide /
ingredient list] stored in the product metafield
[namespace.key] of type [type]. Write the Liquid to render it only when
the metafield exists, with a sensible fallback when it is empty. Match
the existing product template structure I am pasting below.

[paste relevant product template Liquid]

Tip: always ask for the “metafield empty” branch - missing metafields are the most common cause of broken PDPs.

For deeper theme work where one prompt is not enough, see our guide on multi-agent Shopify theme development.

Product description and PDP prompts

Feed Claude your brand voice once at the top of a chat, then run these against each product.

Write a product description with structure.

Write a product description for the product below. Structure it as:
one-sentence hook, two short benefit paragraphs, and a 4-bullet
spec list. Write in [brand voice: e.g. warm, plain-spoken, no hype].
Target customer: [audience]. Do not invent specs that are not in the
data I give you.

Product: [name]
Key facts: [materials, dimensions, use case, differentiator]

Rewrite a description to lead with the outcome.

Rewrite this product description so it leads with the customer outcome
instead of the feature list. Keep it under 90 words. Keep every factual
claim from the original; do not add new claims.

[paste current description]

Generate PDP bullet points that map to objections.

Here is a product and the three most common reasons shoppers hesitate
to buy it: [objection 1], [objection 2], [objection 3]. Write five
benefit bullets for the product page where each bullet quietly answers
one of those objections. Keep each bullet under 12 words.

Product: [paste details]

Draft a size and fit block.

Write a short "size and fit" block for this apparel product page.
Include: how it runs (true to size / size up / size down), the model's
height and the size they wear, and fabric stretch. Use only the data
below. If a field is missing, write "[merchant to confirm]" rather than
guessing.

[paste fit data]

Turn raw specs into a comparison table.

Turn these raw product specs into a clean two-column spec table in
Markdown, with the most purchase-relevant spec first and the least
relevant last. Group related specs together.

[paste specs]

Tip: paste the Markdown table into a metafield and render it on the PDP, so the same source feeds every product.

Write variant-aware copy.

This product has [N] variants: [list them]. Write a single product
description that works across all variants, plus one short variant-
specific line for each that I can show when that variant is selected.
Keep the shared copy neutral about color/size so it never contradicts
the selected variant.

SEO and metafields prompts

Write a title tag and meta description.

Write an SEO title tag (under 60 characters) and a meta description
(under 155 characters) for this Shopify [product / collection] page.
Primary keyword: [keyword]. Lead the title with the keyword. The
description should name a specific benefit, not a generic CTA. Give me
three options for each.

Page topic: [describe]

Generate structured data for a product.

Write valid Product schema JSON-LD for the product below, including
offers with price, currency, and availability. Use Shopify Liquid
variables (product.title, product.price, etc.) so it renders dynamically
in the theme, not hard-coded values. Return a single <script
type="application/ld+json"> block.

Product fields available: [list]

Tip: ask Claude to validate the output mentally against Google’s required and recommended Product properties before returning it.

Plan a metafield schema for a product type.

I sell [product type]. Design a Shopify metafield schema for this
product type: list each metafield as namespace, key, type, and a one-
line description of what it stores and where it shows on the storefront.
Prefer native Shopify metafield types. Flag any field better modeled as
a metaobject.

Write internal-link anchor suggestions.

Here is a new collection page about [topic] and a list of my existing
collection and blog URLs with their titles. Suggest five internal links
I should add to the new page: give the source phrase to link from and
the target URL, and only suggest links that are genuinely relevant.

New page: [summary]
Existing URLs: [paste list]

Audit a page’s on-page SEO from its HTML.

Here is the rendered HTML of a Shopify product page. Audit its on-page
SEO: check for a single H1, heading order, title and meta length, image
alt text, and any obvious keyword gaps for the target keyword "[keyword]".
Return findings as a checklist with pass/fail and a fix for each fail.

[paste HTML]

Cluster keywords into collection and blog topics.

Group this raw keyword list into topic clusters. For each cluster, name
the cluster, recommend whether it should be a collection page or a blog
post, and suggest the primary keyword for that page. Output as a table.

[paste keyword list]

CRO and landing page prompts

Critique a landing page section.

Act as a conversion copywriter. Here is the copy and structure of a
Shopify landing page section. Critique it for clarity, specificity, and
a single clear next action. Then rewrite it. Keep the rewrite the same
length or shorter.

[paste section copy]

Generate hero headline variants.

Write eight hero headline options for a Shopify landing page selling
[product] to [audience]. Mix angles: outcome-led, objection-led,
specificity-led, and curiosity-led. Label each with its angle. Keep
each under 10 words. No exclamation marks.

Draft an above-the-fold block.

Write the above-the-fold copy for a Shopify landing page: eyebrow,
headline, one-sentence subhead, primary button label, and one line of
social proof. The offer is [offer]. The audience is [audience]. The
button should describe the outcome, not say "Shop now".

Build an FAQ that handles purchase objections.

Write six FAQ questions and answers for the product page of [product].
Source the questions from real buying objections: shipping, sizing/fit
or specs, returns, materials/ingredients, and "is this right for me".
Answers should be two to three sentences, specific, and reassuring
without overpromising.

Write a comparison block against alternatives.

Write a fair comparison block for my Shopify landing page comparing
[my product] to [the common alternative the buyer is considering, e.g.
buying the cheaper generic version]. Use a three-column table:
dimension, alternative, mine. Be honest about where the alternative
wins. Keep it credible, not a hit piece.

For more on assembling these into a full page, see our overview of the Shopify page builder workflow.

Analytics and reporting prompts

Claude does not connect to your store, so these prompts work on data you paste or export, and on writing the queries you run elsewhere.

Interpret an exported sales report.

Here is a CSV export of my Shopify orders for the last 90 days. Summarize:
top 5 products by revenue, revenue trend week over week, average order
value, and the share of revenue from returning vs new customers. Call out
the two most actionable findings. Show numbers in a table.

[paste CSV]

Write a ShopifyQL query.

Write a ShopifyQL query for Shopify's Custom Reports that returns
[metric, e.g. net sales] by [dimension, e.g. product title] for the
[timeframe], sorted descending, limited to the top [N]. Explain in one
line what the query returns.

Build a weekly reporting template.

Design a weekly ecommerce reporting template for a Shopify store. List
the 8 to 10 metrics worth tracking weekly, group them into acquisition,
conversion, and retention, and for each give the definition and why it
matters. Output as a table I can paste into a spreadsheet header row.

Diagnose a conversion-rate drop from numbers.

My Shopify conversion rate dropped from [X]% to [Y]% over [timeframe].
Here are the supporting numbers: sessions [..], add-to-carts [..],
reached-checkout [..], purchases [..], by device and by traffic source.
Walk through where in the funnel the drop is concentrated and list the
three most likely causes to investigate first.

[paste funnel numbers]

Tip: give Claude the funnel step counts, not just the headline rate - the drop-off step is the whole answer.

Customer support prompts

Draft a tone-matched reply template.

Write a customer support reply template for this situation:
[e.g. order delayed past the estimated date]. Brand voice: [voice].
Include a placeholder for the order number and one concrete next step.
Keep it warm, under 120 words, and do not promise anything we cannot
guarantee. Give me a version for email and a shorter version for chat.

Build a macro library.

Here are the ten most common support questions we get: [list]. Write a
reusable macro reply for each, in our brand voice [voice], with bracketed
placeholders for anything order-specific. Output as a numbered list so I
can paste them into [Gorgias / Zendesk / Shopify Inbox].

Turn a support thread into a help-doc.

Here is a back-and-forth support conversation that resolved a recurring
issue. Turn it into a clean help-center article: a clear title, a short
intro, numbered steps, and a "still stuck?" line. Remove the customer's
personal details.

[paste thread]

Write a refund or exchange response.

Draft a response to a customer requesting [a refund / an exchange] for
[reason]. Our policy is: [paste policy]. Apply the policy accurately,
keep an empathetic tone, and clearly state what happens next and the
timeframe. Flag if the request falls outside policy so I can decide.

Debugging prompts

Diagnose a Liquid error.

I am getting this error on my Shopify theme: [paste exact error and where
it appears]. Here is the relevant Liquid. Identify the cause, explain why
it happens, and give the corrected code. Do not change unrelated logic.

[paste Liquid]

Fix a layout or CSS bug.

On my Shopify store, [describe the visual bug, e.g. the product grid
breaks to one column on tablet]. Here is the section's HTML and CSS.
Find the cause, explain it, and return only the corrected CSS with a
comment marking what changed.

[paste HTML + CSS]

Debug a cart or AJAX issue.

My Shopify cart [describe behavior, e.g. quantity updates do not refresh
the cart total without a reload]. Here is the JavaScript handling the
cart. Review it against Shopify's Cart AJAX API, find the bug, and return
the fixed script with comments on what changed.

[paste JS]

Trace a slow page.

Here is a [Lighthouse / PageSpeed] report summary for my Shopify product
page. List the issues in priority order by likely impact on Largest
Contentful Paint, and for each give the specific theme-level fix (image
sizing, render-blocking script, preconnect, lazy-loading). Skip anything
I cannot control inside a Shopify theme.

[paste report summary]

Tip: ask Claude to separate “fixable in the theme” from “needs an app or Shopify-side change” so you do not chase dead ends.

Reproduce a bug as a minimal test.

Reduce this Shopify section to the smallest snippet that still reproduces
[the bug], so I can isolate the cause. Explain what you removed and why
the remaining code is enough to trigger it.

[paste section]

Migration and data prompts

Map a CSV to Shopify’s product import format.

Here are the column headers from my old platform's product export. Map
each one to the correct column in Shopify's product CSV import format.
Flag any Shopify required column that has no source and any source column
with no Shopify home. Output as a mapping table.

[paste headers + 2 sample rows]

Transform an export with a formula or script.

I need to transform this product export for Shopify import. The changes:
[e.g. split "Category > Subcategory" into product type, combine size and
color into variant options, strip currency symbols from price]. Write a
[Google Sheets formula / Python script] that does this and explain how to
run it.

[paste sample rows]

Plan URL redirects for a replatform.

I am moving to Shopify and my URL structure is changing. Here are sample
old URLs and their new Shopify equivalents. Infer the pattern, then write
the redirect rules in Shopify's URL Redirects CSV format (Redirect from,
Redirect to). Flag any old URL with no clear new home.

[paste old → new URL pairs]

Audit imported data for problems.

Here is a sample of products after importing into Shopify. Audit for
common import problems: missing or duplicate handles, blank SKUs, broken
HTML in descriptions, variant options that do not line up, and prices
that imported as zero. Return a checklist of what to fix before going
live.

[paste sample]

Where these prompts stop, and what to pair them with

Every prompt above produces text or code. None of them change your store. Claude writes a section; you still have to paste it into your theme, hook up the schema, preview it, and publish. Claude writes a redirect CSV; you still upload it. That last mile is where most of the time goes.

This is the gap Fudge fills. You describe the change in plain English, Fudge writes the Shopify-native code directly into your theme, shows you a live preview, and only publishes when you approve. Because it is built for Shopify rather than a general model you re-brief every session, it follows store best practices by default, learns your brand and preferences as you work, and keeps changes as drafts your team can find, edit together, and schedule to publish. When something needs a human, experienced Shopify developers are on hand. It is the difference between getting code and getting a shipped change. For the bigger picture, our Shopify AI toolkit maps where each tool fits, and the Shopify store editor handles the on-store edits these prompts only describe.

If you also work in the Shopify admin, our Shopify Sidekick prompt cheatsheet covers the admin-side tasks (analytics, discounts, Flow) that Claude cannot reach.

FAQ

Related: AI-first Shopify development and multi-agent Shopify theme development.