--- title: "Announcing Medusa MCP: Build and operate your store with agents" description: "Medusa MCP is a remote MCP server that gives AI agents direct access to your store data on Medusa Cloud — enabling sales reports, automated workflows, and deployment debugging." url: https://medusajs.com/blog/announcing-medusa-mcp last_updated: 2026-10-20T12:36:27Z --- # Announcing Medusa MCP: Build and operate your store with agents _Published 2026-10-20 by Oliver Juhl, Shahed Nasser in Product_ Today, we're announcing the Medusa MCP: bring your store data closer to your agents. You can use the Medusa MCP with your agent to: - Access your data like subscriptions and B2B vendors, or data that lives in third-party services. - Build agentic flows that facilitate sales operations. - Streamline ecommerce data analysis. - Drive actionable points on the spot. - Debug your Medusa Cloud deployments. And much more. The MCP not only gives you access to existing Medusa endpoints, but also to custom endpoints, data, and workflows that you have added for your store. The Medusa MCP is available to all Medusa Cloud users. It currently provides read access only, with write operations coming soon. ## What is the Medusa MCP? The Medusa MCP, available at [https://cloud.medusajs.com/mcp](https://cloud.medusajs.com/mcp), is a remote MCP server that gives agents access to your store data. It doesn't only work with Medusa-specific data like orders and products; it's tailored to your store's data and workflows, allowing you to manage everything out of the box. You can connect multiple agents and MCP clients to the MCP server, such as Claude, Codex, and more, each having access to specific environments of your projects. You can also manage these connections from the Medusa Cloud dashboard. ## How to Connect to the Medusa MCP? You connect to the Medusa MCP based on the MCP client you're using. ### Claude For Claude, you can use the official [Medusa connector](https://claude.ai/directory/medusa). Once you connect it, you'll be asked to log in with Medusa Cloud and choose the environments that Claude can access. ### Claude Code To connect with Claude Code, run the following command in your terminal to add the MCP server: ```bash claude mcp add --transport http medusa-commerce \ https://cloud.medusajs.com/mcp ``` Then, start Claude Code and authenticate with the MCP server: ```bash claude /mcp ``` ### Codex To connect with Codex, run the following commands in your terminal: ```bash codex mcp add medusa-commerce \ --url https://cloud.medusajs.com/mcp codex mcp login medusa-commerce ``` ### ChatGPT Our official ChatGPT plugin is coming soon. In the meantime, you can add the Medusa MCP as a custom plugin. Learn more in [the documentation](https://docs.medusajs.com/cloud/medusa-mcp/chatgpt#add-medusa-mcp-as-a-custom-plugin-in-chatgpt). ### Other Agents and MCP Clients For other agents and MCP clients like Cursor, VSCode, or n8n, refer to [our documentation](https://docs.medusajs.com/cloud/medusa-mcp) for more details on how to connect. ## How to Use the Medusa MCP? After connecting to the Medusa MCP, you can use it in many ways with your MCP client or as part of your agentic flows. ### Work with Your Store's Data All commerce data in your store can be accessed through the Medusa MCP. You don't need to host or configure it separately. Once you deploy to Medusa Cloud, the Medusa MCP allows your connected agent to access this data. For example, if you've integrated a content management system (CMS) like Sanity with Medusa, you can query your Sanity data along with your product data to find gaps in content for a better storefront experience. ### Create Sales Reports with AI Agents You can ask your agent, like Claude, to build you a sales report to share with investors. The agent will retrieve sales data such as orders, promotions, and returns, and create a report based on the trends it finds. ### Schedule AI Agents to Delegate Ecommerce Tasks If your AI agent supports scheduled tasks, you can ask it to run once a week to find actionable points from the week based on store data. For example, the agent can use the Medusa MCP to retrieve products low on stock or orders that still need fulfillment. To make this more efficient, you can integrate task-management tools like Linear as part of this agentic flow, and have your agent automatically create and assign tickets based on your store's data. ### Debug Failed Deployments with AI Agents The Medusa MCP also gives your agent access to your Medusa Cloud project details that are useful for your development and application monitoring. For example, if you build a feature with your AI agent like Claude Code, you can also ask it to debug why the deployment of that feature failed in Medusa Cloud. The agent can use the Medusa MCP to retrieve the logs, ask the Medusa documentation questions for Medusa-relevant errors, and push a fix for your application. This accelerates your development workflow so that you can quickly ship reliable features to your customers. ## Submit Feedback on the Medusa MCP The Medusa MCP has a tool that allows you to submit feedback directly to our team. If you face any issue or need features that are currently missing, you can ask your AI agent to submit that feedback to us. We look through submitted feedback daily and apply fixes promptly. We may reach out to you directly if we need any more details or clarifications. ## Get Started with Medusa Cloud To start using the Medusa MCP, get started with [Medusa Cloud](https://cloud.medusajs.com/) and deploy your project. You can then start managing it with your AI agent. Refer to the documentation for more details on the [Medusa MCP](https://docs.medusajs.com/cloud/medusa-mcp).