Operate with AI · WooCommerce
Connect Claude or ChatGPT to WooCommerce — operate your store with AI
How to give your own AI agent (Claude / ChatGPT) the keys to your WooCommerce store and run it agentically — the method from The AI-Run Store: the right scopes, safe writes, and a human in the loop.
Create a dedicated agent user at the least-privilege role the job needs; issue it a WordPress Application Password (native since WP 5.6).
Give the agent the WooCommerce REST + WordPress REST APIs (catalog, orders, content). For theme/plugin operations, give WP-CLI or SSH.
Put the credential in .env; the agent authenticates with it. The role is the scope — keep it tight.
Rehearse on a WordPress dev instance (VPS / local Docker) you own before pointing the agent at the live store.
- Verify
- Run read-only first and check the result; write only after a backup + preview.
- Multilingual
- WordPress Application Passwords inherit the user's ROLE — there are no per-scope tokens. So create a dedicated agent user at the least-privilege role the job needs (Shop Manager for catalogue, Administrator only if truly required) and issue the application password to that user. Deeper operations (theme, plugins) need WP-CLI/SSH.
Hand this to your own agent
This runbook wires an agent into your store and gives it a working order. It reads first, backs up before every write and leaves the publishing to you.
You are helping me connect my AI agent to my WooCommerce 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 dedicated agent user at the least-privilege role the job needs; issue it a WordPress Application Password (native since WP 5.6). 2. Give the agent the WooCommerce REST + WordPress REST APIs (catalog, orders, content). For theme/plugin operations, give WP-CLI or SSH. 3. Put the credential in .env; the agent authenticates with it. The role is the scope — keep it tight. 4. Rehearse on a WordPress dev instance (VPS / local Docker) you own before pointing the agent at the live store. SCOPES: WordPress Application Passwords inherit the user's ROLE — there are no per-scope tokens. So create a dedicated agent user at the least-privilege role the job needs (Shop Manager for catalogue, Administrator only if truly required) and issue the application password to that user. Deeper operations (theme, plugins) need WP-CLI/SSH. Then operate read-first, back up before writes, batch small, and ask me before anything customer-facing goes live.
The runbook is in English on purpose — agent prompts work most reliably in English. You can still talk to the agent in your own language.
The whole book on operating a store with AI agents.
Rather not do it yourself?Take the courseStep-by-step videos and ready agent prompts.
Rather not do it yourself?Done for youStore Operators sets it up and maintains it.
Rather not do it yourself?Agentic commerceThe customer's own AI buys directly from your 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.
WordPress Application Passwords inherit the user's ROLE — there are no per-scope tokens. So create a dedicated agent user at the least-privilege role the job needs (Shop Manager for catalogue, Administrator only if truly required) and issue the application password to that user. Deeper operations (theme, plugins) need WP-CLI/SSH.
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.