An Agent Skill that walks a developer through every field of the Shopify App Store submission form — section by section — and writes a compliant, copy-paste-ready app-store-listing.md inside the current project directory.
Built to the Agent Skills open specification. Works with Claude Code, Cursor, VS Code Copilot, OpenAI Codex, Goose, Droid, and any other skills-compatible agent.
Preparing a Shopify App Store listing is a multi-day slog: 14 separate form fields, each with its own character limits and Shopify's sometimes-opaque compliance rules. Miss a rule and your submission bounces. Skip the competitive research and your listing converts like wet cardboard.
This skill turns listing prep into a single focused session. It:
- Reads your project — scans
README.md,PRD.md,package.json,shopify.app.toml, anddocs/to auto-extract your app's name, function, target merchant, and differentiators. - Researches competitors — searches the Shopify App Store for your top 3-5 competitors and surfaces positioning gaps you can own.
- Walks 14 sections in order, and for each one:
- Shows the official Shopify field name and character limit
- Generates 2-3 options with a recommended pick and reasoning
- Lets you accept, tweak, or type a custom value
- Enforces compliance rules in real time (blocks "the best", "#1", unverifiable stats, description formatting, etc.)
- Writes the locked-in value to
./app-store-listing.mdwith aCOPY-PASTE READYmarker
- Flags everything it can't do — app icon, screenshots, demo video, privacy policy URL — in a manual checklist at the bottom of the output file so you know exactly what's left.
The 14 sections covered: App Name, Tagline, Introduction, Description, Features, Key Benefits, Integrations, Search Terms, Pricing, Categories & Tags, SEO Title & Meta, Developer Contact, Review Instructions, Install Eligibility.
npx skills add ratul0/shopify-app-listingThis drops the skill into your agent's skills directory (e.g. ~/.claude/skills/shopify-app-listing/ for Claude Code, .agents/skills/shopify-app-listing/ for VS Code).
git clone https://github.com/ratul0/shopify-app-listing.git /tmp/shopify-app-listing
mkdir -p ~/.claude/skills
cp -r /tmp/shopify-app-listing/shopify-app-listing ~/.claude/skills/
rm -rf /tmp/shopify-app-listingThen restart Claude Code (or start a new session). The skill shows up automatically — no registration step needed.
Copy the shopify-app-listing/ directory from this repo into whichever folder your agent scans for skills:
| Agent | Skills directory |
|---|---|
| Claude Code | ~/.claude/skills/ |
| VS Code / Copilot | .agents/skills/ (in your project) |
| Cursor | .cursor/skills/ (in your project) |
| Goose | ~/.config/goose/skills/ |
| Codex | ~/.codex/skills/ |
The skill is self-contained — just the SKILL.md and the references/ folder. No dependencies, no npm install, no compile step.
Inside any Shopify app project, start a session with your agent and say one of:
help me prepare my shopify app store listingwrite my app store listingprepare the listing for submissionhelp me with my app tagline(jumps to that one section)
The skill takes over, asks 4-6 confirmation questions up front, then walks all 14 sections. Output lands at ./app-store-listing.md in whatever directory you started the session in.
Typical session length: 20-40 minutes of focused back-and-forth. You can stop at any point and resume later — the output file is updated after each section is confirmed, so you never lose work.
shopify-app-listing/
├── README.md # (this file)
├── LICENSE # MIT
└── shopify-app-listing/ # The skill itself
├── SKILL.md # Frontmatter + workflow
└── references/
├── section-specs.md # Per-field rules, char limits, forbidden phrases
├── copy-frameworks.md # PAS, AIDA, benefit-first rewriting, merchant voice
└── output-template.md # Shape of the generated ./app-store-listing.md
The skill uses the progressive disclosure pattern from the Agent Skills best practices — SKILL.md is a ~200-line scannable workflow, and the deep rules and examples live in references/ and are loaded only when needed.
These are baked into the workflow and override any user instruction to break them:
- No superlatives: "the best", "the first", "the only", "#1", "world-class", "award-winning", "industry-leading", "guaranteed"
- No unverifiable statistics or user counts
- No Shopify name or logo claims in generated copy
- Hard character limits (regenerated, never truncated mid-sentence)
- No URLs, email addresses, testimonials, or support info inside the description field
- No bold, italic, strikethrough, underline, or lists inside the description field
- No references to other Shopify apps by name
Every borderline generation gets a ⚠️ flag in the output file explaining the risk.
- Does not create app icons, screenshots, or promo videos — just specs and alt text
- Does not submit the listing to Shopify
- Does not validate technical requirements (OAuth, GDPR webhooks, billing API)
- Does not generate translated listings (primary language only)
- Does not replace human judgment — it's a structured assistant, not an autopilot
For the other side of the release process (technical submission checklist, GDPR webhooks, Built-for-Shopify prep), pair this skill with shopify-app-release-assistant or the official Shopify App Store best practices.
Issues and PRs welcome. If you've shipped a listing with this skill and spotted a Shopify rule the compliance layer missed, please file an issue with the exact rule and a link to Shopify's documentation.
Planned improvements:
- Evals folder with golden-output tests for each of the 14 sections
- Optional localization pass (EN → FR/ES/DE/JA)
- Smarter competitive-research that reads Shopify App Store HTML directly instead of relying on Google results
MIT — see LICENSE.
- Built on the Agent Skills open specification
- Rules and constraints derived from the Shopify App Store best practices documentation
- Inspired by Anthropic's shopify-app-release-assistant skill pattern