← All posts

Shopify Skills: One 9.1, Two Under 4 — the Platform-Expertise Gap

2026-11-10

The directory's Shopify coverage is a controlled experiment: one theme skill at 9.1, two integration skills at 3.5 and 4.0, all imported together. The spread tells you exactly what "platform expertise" means — and what it doesn't.

The split

shopify-expert — 9.1 — is the foundation layer: building and debugging Shopify themes, meaning .liquid files, theme.json, and sections. Its dimension line is the strongest in the e-commerce category: trigger 9.0, structure 9.3, content 9.0, engineering 9.0, security 10.0, workflow 8.4.

🛠️ shopify-expert 9.1 Builds and debugs Shopify themes (.liquid files, theme.json, sections), develops custom Shopify a… ⬇ 7

Structure at 9.3 means the Liquid reference material is properly offloaded instead of pasted inline. Engineering at 9.0 means the guidance includes a verification loop — render the section, check the theme editor, confirm the custom section registers. Security at 10 is earned: theme work involves file edits and store credentials, and the skill's handling of both is the strictest in the category.

Then there is the other half. shopify-seo — 3.5 — and shopify-inventory-management — 4.0. Both came from the same import batch. Both describe real, expensive Shopify problems. Both score in the directory's bottom decile.

📣 shopify-seo 3.5 A skill from the GitHub repository nexscope-ai/eCommerce-Skills. ⬇ 2 📣 shopify-inventory-management 4.0 Multi-location inventory — transfers, low stock alerts, demand forecasting, safety stock, ABC ana… ⬇ 1

What the 5.6-point gap is made of

The difference is not domain knowledge. All three skills are about Shopify, and Shopify's documented surface is equally public for themes, SEO, and inventory.

The difference is procedure. The theme skill tells an agent what to do, in what order, with how to verify each step. The integration skills describe what the domain is: what canonical handling means, what safety stock is, why multi-location inventory is hard. Descriptions, not procedures. This is the pattern we documented in the e-commerce bimodal analysis — two strategy skills at 8+, eight how-to files below 5 — and it reproduces here at the single-platform level.

The rubric's workflow dimension makes the separation explicit. The theme skill's workflow scores 8.4: ordered steps, confirmation gates before file edits, a rendered-output check. The integration skills' workflow scores sit below 3. You can read a domain description and nod; you cannot execute it.

Why this matters beyond Shopify

Platform lock-in is where consulting money lives, and the directory's data now says something precise about it: the proprietary mechanics — Liquid's section schema, the theme editor's registration quirks — skill-ify beautifully, because they are concrete procedures with verifiable outcomes. The adjacent domains on the same platform — SEO practice, inventory policy — skill-ify terribly when written by the same authors, because they are judgment domains that happen to sit near the platform.

The template that emerges: when you encode platform expertise, stop at the platform's own mechanics. Shopify theme work deserves a 9.1 skill. Shopify SEO deserves a skill that encodes Shopify's SEO surface — metafields as content infrastructure, structured data through the theme — not a general SEO tutorial wearing a Shopify hat.

The opinion: the next wave of platform skills will be judged by exactly this line. WordPress, Salesforce, and every walled garden after them will produce the same split — deep mechanical skills that earn 9s, and thin adjacent-domain files that drag the average. The platforms with the strangest proprietary idioms will skill-ify best, because that is where procedure is scarce enough to be worth encoding.