Documentation AEM AEM as a Cloud Service Tutorials

AEM MCP Server

Last update: September 22, 2026

Applies to:
Experience Manager as a Cloud Service

CREATED FOR:

Beginner
Leader
User
Developer

The AEM Model Context Protocol (MCP) Server brings AEM into your preferred AI-powered IDE or chat-based application, to streamline and accelerate your AEM work. Describe what you want in natural language instead of writing low-level API code or navigating through the AEM UI.

AEM MCP Server

Today, the AEM MCP Server covers:

Content: Create, read, update, and delete (CRUD) operations for pages and content fragments, plus asset import, search, and management. Supports both read-write and read-only use; access follows the permissions of your signed-in AEM user.
Experience Governance: Evaluate content (text, images, pages) against your brand governance rules, and list brand configurations and checks. Requires an Agents trial or paid license.

TIP

The AEM MCP Server’s tools will improve and grow over time. To see what’s available now, ask your AI AEM MCP what can you do for me?.

The AEM MCP Server section of the Experience League documentation has the full, up-to-date capability list.

Connect your AI tool

The AEM MCP Server can be added to your AI client for seamless integration with various AI tools and platforms. Choose your client below.

AEM MCP Server URL

code language-none
`https://mcp.adobeaemcloud.com/adobe/mcp/aem`

CX Enterprise Coworker

Coworker is Adobe’s conversational AI for CX Enterprise, with native support for the AEM MCP Server.

OpenAI ChatGPT

Install the AEM plugin from the ChatGPT plugin store, or follow the step-by-step setup guide.

Anthropic Claude

Add the AEM connector from Claude’s connector directory, or follow the step-by-step setup guide.

NOTE

Cloud Manager and Cloud Migration aren’t part of /aem yet, so they still need their own registration. The Domain-Specific MCP Servers section has the details.

AEM MCP in action

These examples show how to put the AEM MCP Server to work on real tasks: review and update test content from your IDE, manage and publish live website content by chat, run and debug a Cloud Manager pipeline, check a page against your brand guidelines, or search a Content AI index. Each one walks through a specific scenario end to end: connect the server in your tool of choice, then get a concrete result back.