Contentful Delivery MCP Server

What is Contentful Delivery MCP Server?

The Contentful Delivery MCP Server is an MCP (Model Context Protocol) server that provides seamless access to Contentful's Delivery API through AI assistants, enabling querying and retrieval of content entries, assets, and content types using natural language.

How to use Contentful Delivery MCP Server?

Install via npm (npm install @mshaaban0/contentful-delivery-mcp-server or globally) and set environment variables CONTENTFUL_SPACE_ID and CONTENTFUL_ACCESS_TOKEN. Optionally restrict content by setting CONTENTFUL_CONTENT_TYPE_IDS. Invoke the server via npx @mshaaban0/contentful-delivery-mcp-server or integrate with Mastra AI using the provided TypeScript example.

Key features of Contentful Delivery MCP Server

• Natural language queries to search content.
• Retrieve entries by ID or content type.
• Asset management and asset details retrieval.
• Content type schema access.
• Pagination support for large datasets.
• Rich text content handling.

Use cases of Contentful Delivery MCP Server

• Query articles about machine learning in natural language.
• Fetch a specific blog post by its entry ID.
• List all assets (images, files) stored in Contentful.
• View the content type schema for custom models.
• Integrate Contentful content with AI agents in Mastra.

FAQ from Contentful Delivery MCP Server

What dependencies are required?

Node.js and npm are required. You also need a Contentful account with a Space ID and a Delivery Access Token.

Where does the data live?

Data is stored in Contentful and accessed via the Contentful Delivery API; the server does not store data locally.

What authentication is needed?

Set the CONTENTFUL_ACCESS_TOKEN environment variable with your Contentful Delivery API token.

Can I restrict which content types are accessible?

Yes, set the optional CONTENTFUL_CONTENT_TYPE_IDS environment variable to a comma-separated list of content type IDs.

What transport does this server use?

The server uses STDIO transport (command-line invocation) and is intended for use with MCP‑compatible clients.