Claude operating calendar for Shopify brands (prompts included)

I've been building prompt libraries for running Shopify operations in Claude.

Most either skip audit-related prompts entirely or run them at the wrong time. Checking something monthly that needs a weekly look, or letting one slide until a number moves and there's no early warning left to catch it.

I am sharing 5 prompts, useful whether you're running growth, ops, or lifecycle at a brand - from different playbooks, mapped to the cadence they're actually built for.

→ One prompt you run daily
→ One you run weekly
→ One you run monthly
→ Two you run quarterly

1/ Daily orders briefing

The usual morning routine is to open a sales dashboard and click through two or three tabs to piece together what happened yesterday. This Claude prompt replaces that with one message that answers the same question in three sentences.

→ Total sales, order count, and average order value for yesterday specifically, not a rolling average that can mask a bad day
→ Top 5 products by units sold, which can drift from what you'd assume is selling well
→ Revenue split by where the order came from: direct, email, paid, organic
→ Any order with a failed payment or an open refund request, flagged separately

The prompt:
"Pull yesterday's orders. Use run-analytics-query. Compute: 1. FROM sales SHOW total_sales, orders, average_order_value SINCE yesterday UNTIL yesterday 2. FROM sales SHOW units GROUP BY product_title ORDER BY units DESC LIMIT 5 SINCE yesterday 3. FROM sales SHOW orders GROUP BY order_referrer_source SINCE yesterday. Flag any orders with payment_status = failed or open refund requests. Format the output as 3 sentences I can read in 30 seconds."

From 28 Prompts for Shopify's New Agentic Guidelines.

The other 25 prompts in that book cover the diagnostic and catalog-enrichment work that has to happen before the weekly review below is even worth running, plus a non-US playbook for stores selling outside the US and Canada.

2/ Weekly agentic readiness review

Shopify auto-indexed every eligible store into the agent network, so ChatGPT, Copilot, Gemini, and Perplexity are already reading your catalog and ranking it on their own logic, whether or not you've touched anything.

This is the Monday report, but for AI visibility instead of sales.

→ Listing Quality trend: products improved, regressed, or stayed flat over the last 7 days
→ Metafield fill gap: how many products still have blank fields on material, style, occasion, the data an AI agent needs to recommend that product at all
→ Direct checkout health: any product newly ineligible for one-click checkout inside ChatGPT or Copilot
→ Revenue by agentic channel: ChatGPT, Copilot, Gemini, Perplexity, Shop
→ Get one ranked action item for the week in the end, not a list of ten

The prompt:
"Run my weekly agentic readiness review. Report: 1. Listing Quality trend: how many products improved, regressed, stayed flat in the last 7 days. 2. Metafield fill gap: percentage filled across material, style, occasion, tech_specs, top_features, USPs (vs last week). 3. Direct checkout health: any products newly flagged as ineligible (subscriptions, bundles, customizable). 4. Agentic channel performance: orders and revenue per channel (ChatGPT, Copilot, Gemini, Perplexity, Shop) from the Agentic admin section. 5. Top action item: the single most leveraged fix this week."

Want to check without Claude? Take a free audit here - runs the same Listing-Quality-style read on your store for free, no setup or login required.

Full playbook here: 28 Prompts for Shopify's New Agentic Guidelines →

That covers the store side. The next two shift over to the inbox, specifically Klaviyo, and to the parts of email performance that don't show up until you look deeper.

3/ Monthly performance report (Klaviyo)

The report that goes to the founder, not the one you read alone. A single week of Klaviyo data is noisy enough that the real pattern usually only shows up over a full month.

→ Headline revenue, this month against last month and against the same month last year
→ Channel split between email and SMS
→ Campaign performance, top 5 plus the average
→ Flow performance, ranked by revenue contribution
→ List growth, adds and unsubscribes together
→ Deliverability signals folded into the same report
→ Closes with one written paragraph on what to prioritize next month, the only part not formatted as a table

Month-over-month alone misses seasonality. Year-over-year alone misses momentum. Run together in the same report, they cover for each other.

From 27 Prompts for Klaviyo in Claude. The rest of that book covers weekly reporting, flow audits, segment cleanup, campaign pre-send checks - day-to-day prompts.

4/ Quarterly deliverability and consent audit (Klaviyo)

The one almost nobody runs until a send underperforms and they go looking for why, by which point the damage has usually been building for months already.

→ Bounce rate by domain: Gmail, Yahoo, Outlook, iCloud
→ Spam complaint rate by campaign type
→ Open, click, and unsub trends by week, not a single quarterly average
→ Consent source for every subscriber: form, checkout, manual, import, the record you'd want on hand if a list were ever challenged
→ Sunset candidates, no engagement in 6 months
→ Suppression list health
→ SPF, DKIM, and DMARC status in one pass

Deliverability is the one metric on this whole list where doing nothing actively makes it worse. The list you have right now is the best it will ever be without upkeep.

5/ Quarterly competitor drift tracker (Meta ads)

This one turns the same idea outward, toward whichever competitor you keep an eye on. Instead of a typed prompt, you paste in two sets of screenshots from the competitor's Meta ad library, taken about 90 days apart.

→ Visual direction and copy voice, compared across the two dates
→ Which ads they're still scaling versus which ones they've dropped
→ A suggestion for what to test next, based on that shift

No connector - works directly in Claude.

The other five prompts in that book, /audit, /winners, /voice, /gap, /brief, are the full workflow - this one only samples once a quarter.

Everything above assumes the catalog underneath these prompts is already readable, real attributes filled in, not blanks an agent has to guess at.

That's the layer we built Atomz for.

We call it the Agent OS for a Shopify store.

Catalog Genius maps every item to the Shopify taxonomy and writes the structured attributes. AI Search and AI Assistant run natural-language discovery on top of that same catalog, on your store and off it. AI Insights, still in early access, will be showing the agent-readiness score and what each gap is costing you by channel.

Try for free on Shopify →

- Ankit