Shopify AI Toolkit
Shopify's official agent plugin: docs and API schema search, GraphQL and Liquid validation, and store management
Install
claude plugin install shopify-ai-toolkit@claude-plugins-officialThis is third-party code. Review the repository files before installing.
What it does
The plugin connects an agent to the Shopify platform. The agent searches docs and API schemas without leaving the editor and validates GraphQL queries, Liquid templates and UI extensions against Shopify schemas. Through Shopify CLI capabilities it can run operations on a store. The plugin installs into Claude Code, Codex, Cursor, VS Code and several other agents and updates automatically. Where plugins are not supported, skills or the Dev MCP server can be installed separately per shopify.dev docs.
Who it is for. For developers building Shopify apps, themes and extensions.
Good fit when
- The agent writes Admin API GraphQL queries and must follow the current schema
- You need to validate a theme's Liquid template before deploying
- You want the agent to run an operation on your store
Not a fit when
- Your store is not on Shopify
- Queries and validated code must not reach telemetry and you cannot opt out
Example request
Write an Admin API GraphQL query for the last 20 orders and validate it against the schemaLimitations
Telemetry is on by default: scripts send search queries, validated code and, in Claude Code, the user's latest message to Shopify. Opt out with ~/.config/shopify-ai-toolkit/opt-out or OPT_OUT_INSTRUMENTATION=true. Store management needs a Shopify account. Shopify does not serve merchants in Russia.
How to disable. In Claude Code run claude plugin uninstall shopify-ai-toolkit; in other agents remove the plugin with their tools. The opt-out file turns off telemetry.
Security check
- Runs store operations through Shopify CLI
- Sends queries, code and user messages to Shopify by default
README in short
The README gives install commands for Claude Code, Codex, Antigravity, Cursor, Hermes, OpenClaw, Pi and VS Code. It briefly lists capabilities: docs and schemas, code validation, store management and auto-updates. A large section details telemetry, hooks and how to opt out. For platforms without plugins it points to shopify.dev docs.
FAQ
What exactly goes to telemetry?
Tool and skill names, search text and response, validated code and its context, and in Claude Code the user's latest message truncated to 2000 characters when a Shopify skill fires.
Why does the env var not always disable telemetry?
Some agents spawn scripts and MCP processes without your environment. The opt-out file is more reliable.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
GitHub toolkit for spec-driven development: the specify CLI adds agent commands and skills to a project, from principles to implementation
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything