Connect Claude or ChatGPT to Shopify — operate your store with AI
How to give your own AI agent (Claude / ChatGPT) the keys to your Shopify store and run it agentically — the method from The AI-Run Store: the right scopes, safe writes, and a human in the loop.

For a developer
- Create a Custom Distribution app (Partner/Dev Dashboard or CLI) and install it on the store — the in-admin 'Develop apps' path is blocked as of Jan 2026.
- Grant broad Admin scopes at install (all except read_all_orders) so your agent can run products, inventory, content, themes and translations without a reinstall later.
- Fetch the Admin API access token via client_credentials (Client ID + API secret → /admin/oauth/access_token; ~24h, refresh). Put it in .env as SHOPIFY_ACCESS_TOKEN.
- Point your AI agent (Claude / ChatGPT) at the store: it authenticates with that token against the Admin GraphQL API (and the Theme Asset API for theme edits). MCP is a transport, not a credential — you still mint this token.
For real operations, grant broad Admin scopes at install — in practice all Admin scopes EXCEPT read_all_orders. Re-scoping later means a new app version + store re-approval, so plan ahead. read_all_orders needs Shopify's special approval (without it you get ~60 days of order history). Apps touching protected customer data must be approved by the store OWNER, not a staff admin.
Run read-only first and check the result; write only after a backup + preview.
The exact code

Prompt for your AI agent — connect Shopify & implement
A complete copy-paste prompt for your own AI agent (Claude / ChatGPT). It tells the agent how to connect to your Shopify store and implement the JSON-LD for you — with safe-write discipline: backup, preview, verify, and your approval before publishing.
You are helping me connect my AI agent to my Shopify store so it can operate the store for me, safely. SAFE-OPERATING DISCIPLINE (follow in order): 1. Read first: point the agent at read-only tasks and confirm it sees your real data before any write. 2. Back up before every write; rehearse unfamiliar jobs on a dev/staging store first. 3. Plan the ring of keys for what you'll let the agent do over the coming months (re-scoping is painful on some platforms), and grant that at connect time. 4. Batch writes in small clusters, not one giant run; respect the platform's rate meter. 5. Store the credential in .env, reference it by name, never print it; keep it revocable from a UI you own. CONNECT: 1. Create a Custom Distribution app (Partner/Dev Dashboard or CLI) and install it on the store — the in-admin 'Develop apps' path is blocked as of Jan 2026. 2. Grant broad Admin scopes at install (all except read_all_orders) so your agent can run products, inventory, content, themes and translations without a reinstall later. 3. Fetch the Admin API access token via client_credentials (Client ID + API secret → /admin/oauth/access_token; ~24h, refresh). Put it in .env as SHOPIFY_ACCESS_TOKEN. 4. Point your AI agent (Claude / ChatGPT) at the store: it authenticates with that token against the Admin GraphQL API (and the Theme Asset API for theme edits). MCP is a transport, not a credential — you still mint this token. SCOPES: For real operations, grant broad Admin scopes at install — in practice all Admin scopes EXCEPT read_all_orders. Re-scoping later means a new app version + store re-approval, so plan ahead. read_all_orders needs Shopify's special approval (without it you get ~60 days of order history). Apps touching protected customer data must be approved by the store OWNER, not a staff admin. Then operate read-first, back up before writes, batch small, and ask me before anything customer-facing goes live.
Questions people ask their AI
- What is the prompt to connect Claude to Shopify to run my store?
- Use the runbook on this page: mint a scoped credential, load it from .env, and tell Claude to authenticate against the store's API, read first, back up before writes, and ask you before publishing.
- Which scopes does an AI agent need to operate a Shopify store?
- For real operations, grant broad Admin scopes at install — in practice all Admin scopes EXCEPT read_all_orders. Re-scoping later means a new app version + store re-approval, so plan ahead. read_all_orders needs Shopify's special approval (without it you get ~60 days of order history). Apps touching protected customer data must be approved by the store OWNER, not a staff admin.
- Is it safe to let an AI agent operate my live store?
- Yes, with the discipline in The AI-Run Store: read-first, back up before every write, rehearse on a dev store, batch small, and keep a human approving anything customer-facing. This page's runbook builds those in.
Sources
- The AI-Run Store — ch. 6 'Wiring in' (giving agents the keys to your platform)
- Legacy custom apps can't be created after Jan 1 2026 (Shopify changelog)
- Shopify Admin API — access tokens (incl. client_credentials)
- Shopify themeFilesUpsert (theme file writes; needs write_themes_assets)
- Shopify — manage translated content (metafields via translationsRegister)
- Shopify Liquid translate filter (only locale-file strings translate)
Get found and recommended by AI
Store Operators makes your store AI-visible and agent-ready — from a free audit to a done-for-you data layer and agentic commerce.