I run a small Shopify store with my partner, and every morning I opened four admin screens to answer one question: what changed overnight? Shopify's own AI toolkit gives an agent the rails — Admin GraphQL, ShopifyQL, Liquid. What it doesn't give you is a routine. The agent still does only what you prompt it to do, from scratch, every time. StoreHand is the routine on top: which questions get asked, which thresholds apply to your store, in what shape the answer arrives. Where it stands today, five days in: 74 commits, 121 tests, four of the six v1 skills shipped. ✅ storehand-setup — profile and connection ✅ daily-store-briefing — orders, payments, cancellations, stock ✅ store-health-check — dead links, discount windows, SEO gaps ✅ product-listing-writer — copy you approve before it's written ⏳ price-and-competitor-watch ⏳ seo-metadata-audit ⏳ weekly-store-report One rule keeps the pace honest: a skill is only listed as shipped after it has run against a live store, and the evidence goes in the repo — the commands, the real output, and what was awkward about it. Building the listing writer that way turned up nine real bugs before it shipped. A typed line of tildes silently truncated a description. Any product with more than one photo was rejected outright. An empty SEO field conflicted with itself, because Shopify returns null where the proposal recorded an empty string. None were visible until it ran against a real shop. All of it is written up, including what I got wrong. And the six aren't the finish line. They're what a store needs on day one. The roadmap gets updated as my own store keeps finding gaps — returns, restock timing, supplier alt text that reads like a wholesale catalogue. That last one is already logged as the next thing to fix. If you run a Shopify store from a terminal, I'd like you to break this. Especially the proposes-you-approve model — that's the part I'd most like to be argued with about.

Forum Threads

p/storehand

•

3mo ago

Most Shopify apps are 200 lines of Liquid with a subscription attached. Am I wrong?

I've been working on StoreHand an open-source set of Claude Code skills that helps Shopify store owners with everyday ops (morning briefings, stock alerts, product copy, health checks).

We shared it here recently.

Since then, I've been talking with a bunch of store owners, and one topic keeps coming up. Surprisingly, it's not really about operations. It's about the amount of apps they have to rely on.

View all

StoreHand

Claude Code skills that run your Shopify store daily

Claude Code skills that run your Shopify store daily

Forum Threads

Most Shopify apps are 200 lines of Liquid with a subscription attached. Am I wrong?

Forum Threads

Most Shopify apps are 200 lines of Liquid with a subscription attached. Am I wrong?