Most AI tools give developers generic code suggestions. But when your AI agent knows your real content structure, your sites, your content types and fields, your workflows, it will build something that actually works.
That’s exactly what the dotCMS MCP Server does for your dev workflows. It gives your agents the context they need to generate real components, automate migrations, and even pair with design tools.
What Is an MCP Server?
An MCP, short for Model Context Protocol, is a secure connector that lets AI agents and system to interact with any system the same way a human user would.
It gives your AI tools the context they need, for example:
What data exists
How that data is structured
What actions they can perform
What rules and permissions they must follow
So instead of hallucinating APIs or guessing props, your agent works with your real dotCMS data.
How to Set Up the MCP Server
Here’s how to get started.
Get a dotCMS environment Cloud, on-prem, or our demo site — they all work.
Create an AI User and API Token
In your dotCMS admin, create a user just for your AI agent.
Give it only the permissions it needs.
Generate a secure API token.
Configure the MCP Server in your AI IDE (like Cursor, VSCode, Claude Code or Windsurf), add the server following this configuration:
{
"mcpServers": {
"dotcms": {
"command": "npx",
"args": ["-y", "@dotcms/mcp-server"],
"env": {
"DOTCMS_URL": "https://your-dotcms-instance.com",
"AUTH_TOKEN": "your-auth-token"
}
}
}
}Check the MCP server is running and start prompting.
What You Can Do Right Now
Here’s what developers can do with the dotCMS MCP Server today, and why it’s better than generic AI code completions.
1. Generate UI Components That Actually Work
Need a card, list, or a banner component? Your agent already knows your Product, Banner, or Blog content type structure. Ask it to generate a React or Angular component that:
Uses the right props and types.
Pulls in real fields
Includes prop validation or TypeScript interfaces.
Example prompt: “Generate a React component for our Event content type. Use Tailwind CSS. Include prop types and an example of how to pass props.”
2. Migrate Faster
Moving your frontend from JavaScript to TypeScript? Your agent can:
Use your dotCMS schemas to create accurate interfaces.
Rewrite components with strict type checks.
Flag mismatches between old props and your schema.
3. Connect Design + Content
Pair the dotCMS MCP Server with other MCP Servers, like Figma or Storybook, to tie your content structure to your design system.
Your content agent knows the schema.
Your design agent knows the layout.
Your code agent builds a component that does both.
This cuts back-and-forth between dev and design teams — and keeps everything consistent.
4. Automate Checks and Audits
Use your AI agent to:
Audit your component library for missing accessibility props, alt text, or ARIA labels — mapped to real schema fields.
Flag props you’re not using anymore.
Suggest test cases for common content variations.
Ready to Build?
Connect your dotCMS MCP Server. Let your agent learn your schemas. Start with one real task — like generating a component or rewriting a page in TypeScript. See how real context beats generic snippets every time.
Learn how to set up your MCP Server for developers and get started.