Kontent.ai MCP Server provides a bridge that implements the Model Context Protocol, enabling AI models to understand and interact with your Kontent.ai projects. It exposes a suite of patch, management, and content-item operations that can be driven from AI assistants or editor plugins, allowing you to inspect content models, retrieve and modify content types, items, taxonomies, and more using natural language prompts. Available tools cover content-type management, content-item operations, theme-like patch commands, and various retrieval utilities, enabling you to model, query, and evolve your Kontent.ai content structures with AI guidance. To begin, install and run the MCP server, then connect your AI tool (such as Claude, Cursor, or a VS Code integration) to issue the defined commands and receive structured responses that reflect Kontent.ai’s content model and state.

The server supports two transport modes: STDIO for local, interactive use and Streamable HTTP (shttp) for remote or browser-based tooling. Use the corresponding transport when starting the server with the npx command described in Quickstart. Once running, you can invoke tools like get-type-mapi, list-content-types-mapi, add-content-type-mapi, patch-content-type-mapi, and many others to manage content types, snippets, taxonomies, and content items via natural language instructions.

Prerequisites:

• Node.js (v14+ recommended) with npm
• A Kontent.ai account, project, and a Management API key/environment ID as described in the Quickstart

Installation steps:

1. Install Node.js and npm from https://nodejs.org/
2. Install and run the MCP server using npx (no local install required):

Code:

Using STDIO transport

npx @kontent-ai/mcp-server@latest stdio

Alternatively, using the Streamable HTTP transport

npx @kontent-ai/mcp-server@latest shttp

3. If you prefer a local install workflow, you can install the package locally in a project:

Code:

Initialize a new project (if you don’t have one)

npm init -y

Install the MCP server package locally

npm install @kontent-ai/mcp-server@latest

Run it from node_modules/.bin or via npx in scripts

npx @kontent-ai/mcp-server@latest stdio

4. Configure environment variables as needed (see additional notes) and connect your AI tools to the running MCP server using the supported transports.

Notes and tips:

• The MCP server is designed to work with AI tooling that can emit natural language prompts mapped to the available tools (e.g., get-type-mapi, add-content-type-mapi, patch-content-type-mapi).
• Ensure your Kontent.ai Management API key has the necessary permissions for the operations you intend to perform (read/write access to content types, items, taxonomies, etc.).
• When using the patch operations, follow the provided guide (get-patch-guide) for correct operation shapes (move, addInto, remove, replace).
• If you run into connectivity issues, verify that the transport (stdio or shttp) is correctly configured and that your AI tool knows how to reach the MCP server’s endpoint.
• For debugging, use the development-related sections in the project (Local Installation, Project Structure, Debugging) to understand how requests are formed and routed.
• Environment variables related to Kontent.ai credentials and environment IDs may be required; keep credentials secure and avoid hard-coding them in public repositories.