Today we're thrilled to unveil a new way of interacting with our flagship vector database, Astra DB. Say hello to Astra DB over MCP—an innovative way to communicate with your database that leverages the Model Context Protocol (MCP) to let you create and manage databases without writing a single line of code.

What is Model Context Protocol (MCP)?

MCP is an innovation first pioneered by Anthropic in late 2024. It’s a standardized protocol designed for sharing context between language models and tools. This means that any MCP server can communicate with any MCP client, enabling language models to execute functions agentically on your behalf. Imagine being able to hand off entire functions to an AI—MCP makes that possible.

For example, popular MCP clients include:

• Claude Desktop

• Cursor

Both can consume data from MCP servers and act agentically as a result. Let’s get hands on with Astra DB over MCP.

Hands-on with Astra DB over MCP

In our demo, we explore how Astra DB over MCP unlocks a new way of interacting with your data. Let’s walk through the process.

1. Set up your Astra DB environment

To get started, you need an Astra DB application token and an API endpoint. To get these, you’ll have to:

• Sign up for Astra DB - It’s free and quick. Just sign up, create a database, and you’ll receive your API endpoint along with an application token. Here are more detailed instructions to do so.
• Create your database - For this demo, we set up a vector database named “my_mcp_db.” Astra DB’s multi-cloud capability means you can choose your preferred region, and the database is ready within minutes.

2. Integrating with an MCP client

Once your Astra DB instance is ready, you can integrate it with an MCP client like Claude Desktop:

• Configure Claude Desktop - Open the app, go to Preferences → Developer → Edit Config. This will take you to a JSON file. Paste the following JSON configuration snippet that includes your DB token and API endpoint.

{
  "mcpServers": {
    "astra-db-mcp": {
      "command": "npx",
      "args": ["-y", "@datastax/astra-db-mcp"],
      "env": {
        "ASTRA_DB_APPLICATION_TOKEN": "your_astra_db_token",
        "ASTRA_DB_API_ENDPOINT": "your_astra_db_endpoint"
      }
    }
  }
}

• Launch and verify - Restart Claude Desktop and watch as it connects to Astra DB—instantly revealing 10 available MCP tools.

From here, you can ask Claude to do anything you like inside your database: create collections, insert data, clean up, and more. This is a handy way of interacting with your database via an AI assistant, but we can do more when we use Cursor as our MCP client.

Building a full, end-to-end application with a UI, database, and API

The real magic happens when you use Astra DB over MCP in Cursor. To set this up:

1. Go to Settings -> Cursor Settings -> MCP

2. From there, you can add the server by clicking the "+ Add New MCP Server" button and entering the following values:

• • Name - Whatever you want
  • Type - Command
  • Command -

env ASTRA_DB_APPLICATION_TOKEN=your_astra_db_token ASTRA_DB_API_ENDPOINT=your_astra_db_endpoint npx -y @datastax/astra-db-mcp

Once added, your editor will be fully connected to your Astra DB database.

Now you can invoke the Cursor agent (by pressing Cmd+I on macOS) and ask it to build anything you want: whenever a database is needed, it will automatically operate Astra DB to do whatever is required.

In our demo in the video above, the language model agent executes a series of tasks: from setting up the collection to auto-generating Next.js route handlers and fixing UI issues on the fly. The result? A fully functional to-do list app powered entirely by Astra DB over MCP.

Why this matters

Astra DB over MCP demonstrates the incredible potential of combining any tool with AI agents. By enabling agentic interactions between tools and language models:

• Developers can accelerate time to production without the overhead of boilerplate code.

• Non-technical users can create applications that are normally reserved for seasoned programmers.

• Innovation is democratized, letting you build everything from a Twitter clone to a YouTube replica with minimal effort.

What’s next?

We’re excited to see what you’ll build using this new mode of development. Whether you’re a developer, a startup founder, or a tech enthusiast, Astra DB over MCP opens up a world of possibilities. So, what will you create? Join the conversation on Discord, try out Astra DB over MCP, and let us know how you’re leveraging the power of agentic database interactions.

---

Frequently Asked Questions (FAQ)

1. What is Astra DB over MCP?

Astra DB over MCP is a new method of interacting with our flagship vector database, Astra DB, using the Model Context Protocol. It allows you to perform database operations through prompts with an AI agent—without writing any code.

2. What is the Model Context Protocol (MCP)?

MCP is an open standard, first pioneered by Anthropic in late 2024, that enables seamless communication between language models and external tools. It allows AI systems to share context and execute functions agentically on your behalf.

3. Do I need to write any code to interact with Astra DB over MCP?

No! One of the key benefits of this new integration is that you can perform complex database operations—such as creating collections, inserting data, and building entire applications—without writing a single line of code. MCP clients like Claude Desktop, Cursor, etc. manage all the interactions for you.

4. How do I get started with Astra DB over MCP?

Simply sign up for Astra DB, create your database to receive an API endpoint and an application token, and then configure your MCP client by updating its settings with these credentials.

5. What are some examples of MCP clients?

Popular MCP clients include Claude Desktop—a desktop application for interacting with AI models—and Cursor, an AI-enabled version of VS Code that integrates MCP tools directly into your development workflow.

6. Is Astra DB over MCP an open-source project?

Yes, AstraDB over MCP is an open-source project. You can access the code and contribute to its development via GitHub.

8. Where can I get help if I encounter issues?

You can refer to our detailed documentation, check out the GitHub repository for troubleshooting tips, or join our Discord community where fellow users and developers share advice and best practices.