MCP server

The Emfas MCP server lets you use your own AI assistant with your product catalog. Connect an assistant such as Claude or Cursor, and ask it to find products, check data quality or prepare new content, in the same chat you already use.

MCP (Model Context Protocol) is an open standard that AI assistants use to connect to other tools. You do not need to know how it works to use it.

What you can do

The assistant gets the same tools as the Emfas agent. For example, you can ask it to:

Find and read products - "Which jackets are missing a care description?"
Check completeness - "Show me the completeness report for the Spring collection"
Prepare content - "Write SEO titles for all sneakers and put them in a proposal"
Translate - "Translate the product descriptions in this proposal to German and French"
Manage your setup - attributes, families, instructions, completeness profiles, languages, feeds and automations
Search the Emfas docs - "How do formulas work in Emfas?"

You stay in control

Changes to product content go into a proposal. You can review every suggestion in Emfas before you publish it. The assistant only publishes when you tell it to.

Server address

Use this address when your assistant asks for the server URL:

https://api.emfas.ai/mcp

The first time you connect, a browser window opens. Sign in with your Emfas account and choose the workspace to connect. The assistant can then only see and change that workspace.

Connect your assistant

Claude

Open Settings in Claude and go to Connectors.
Click Add custom connector.
Enter Emfas as the name and https://api.emfas.ai/mcp as the URL.
Click Connect and sign in to Emfas.

Claude for teams

On a Team or Enterprise plan, an owner adds the connector once for the organization. Each person then connects with their own Emfas account.

Claude Code

Run this command in your terminal:

claude mcp add --transport http emfas https://api.emfas.ai/mcp

Then type /mcp in Claude Code and select emfas to sign in.

Cursor

Add Emfas to your mcp.json file:

{
  "mcpServers": {
    "emfas": {
      "url": "https://api.emfas.ai/mcp"
    }
  }
}

Cursor asks you to sign in the first time it uses the server.

Other assistants

Any assistant that supports remote MCP servers can connect. Look for a setting called "connectors", "integrations" or "MCP servers", and use the server address above.

Connect with an API key

Scripts and assistants that cannot open a sign-in window can use an API key instead. Send the key in the Authorization header:

Authorization: Bearer emf_at_...

The key gives access to the workspace it was created in.

Keep your key secret

Anyone with the key can read and change your catalog. Do not put it in shared files or chats.

Tips

Be specific - name the products, attributes and languages you want. "Rewrite the description for all products in the Outerwear family" works better than "improve my products".
Review before you publish - open the proposal in Emfas to check the suggestions, then publish from there or ask the assistant to do it.
Use your guidelines - the assistant follows the guidelines and instructions you set up in Emfas, the same as the Emfas agent.

MCP server

On this page