Talk to Your Shopify Inventory in Claude, ChatGPT, or Cursor

You probably already have an AI assistant you trust. Maybe it is Claude, maybe it is ChatGPT, maybe it is Cursor or VS Code while you are building something. You use it every day. The one thing it has never been able to do is see your actual inventory.

Most inventory apps that add AI keep it locked inside their own interface. You get a chat box in their dashboard, and that is where the intelligence stays. Stockful has an in-app assistant too, and it is genuinely useful. But your inventory data should not be trapped in a single window.

Stockful now runs a remote MCP server. In one step you can connect your inventory data to the AI tools you already use, and ask about your stock, forecasts, margins, and anomalies in the same place you do the rest of your work.

What an MCP server is, briefly

MCP stands for Model Context Protocol. It is an open standard for connecting AI assistants to outside tools and data sources in a controlled way. Think of it as a universal adapter. Instead of every app building a bespoke integration for every AI tool, an app exposes one MCP server, and any tool that speaks MCP can connect to it.

A growing list of tools support it, including the Claude desktop app, Claude Code, Cursor, VS Code, the Gemini CLI, and ChatGPT. Stockful's server is remote, which means there is nothing to install or host. You connect to it over a single URL.

Ask about your inventory from your own AI tool

Once connected, your AI tool can read your real Stockful data. That covers:

• Current inventory levels across your locations
• Demand forecasts and reorder timing
• Stock value and margins
• ABC ranking, so you can see which SKUs actually drive revenue
• Lost sales from stockouts
• AI-detected demand anomalies, the unusual spikes and drops Stockful flags automatically

Because this runs inside the tool you already use, you can fold inventory into whatever you were already doing. Ask Claude to draft a supplier email and pull the reorder quantities from Stockful in the same conversation. Ask ChatGPT to summarise the week's anomalies before your Monday planning. The inventory data sits alongside everything else rather than in a separate tab you have to remember to open.

Take action, with a confirmation step

The connection is not limited to reading. You can also make changes through it, with a confirmation before anything actually happens. Through your AI tool you can:

• Snooze reorder recommendations
• Dismiss transfer suggestions
• Toggle alerts and scheduled reports on or off
• Adjust forecast settings

Each write action asks for confirmation first. The assistant proposes, you approve. Nothing changes in your store on the assistant's say-so alone.

Setup: one URL, no keys to manage

Connecting is deliberately simple. In Stockful, go to Settings, then MCP server. Create a connection, give it a name so you can recognise it later, and copy the single URL it generates. Paste that URL into your AI tool's connector settings. There are no API keys to generate and no headers to configure.

You create one connection per tool, which keeps things tidy and easy to manage. Full setup steps are in the MCP server documentation.

Built with the right boundaries

Letting an AI tool reach into your business data needs clear limits. Here is how each connection is scoped:

• Read-only or confirmed-write. A connection either reads your data, or for the write actions above, asks for confirmation before it changes anything.
• Scoped to your store. A connection only ever sees your shop's data.
• Attributed. Each connection is tied to whoever created it, so you have a record of who connected what.
• Revocable any time. If you no longer want a tool connected, revoke the connection and access stops immediately.

As with the rest of Stockful's AI features, your shop data is sent to AI models for inference only. It is not used to train models.

How this fits with the in-app assistant

Stockful already has an AI assistant built into the app, so why also expose an MCP server? Because they suit different moments.

The in-app assistant is there when you are working inside Stockful and want answers without leaving your reports. The MCP server is for when you are working somewhere else, in Claude, ChatGPT, or your editor, and you want your inventory to come to you. Same underlying data, two ways in. You do not have to choose one.

Getting started

Three steps to your first answer:

1. In Stockful, open Settings, then MCP server, and create a connection.
2. Copy the URL and paste it into your AI tool's connector settings.
3. Ask it something real, like "what should I reorder this week?" and check that the answer references your actual variants and quantities.

Your inventory data has always held the answers. Now you can ask for them without changing tools.

Get started free at stockful.app. Explore the full AI feature set at stockful.app/ai.

More from Stockful:

• Stockful AI: an inventory assistant that actually knows your Shopify store
• Move stock instead of reordering it: actionable transfers and incoming stock
• Why Shopify inventory management needs more than native tools