Netlify MCP Server: AI Agents Can Now Deploy Code Directly

Netlify MCP Server is a part of Agent Week. A week of announcements focused on making it easier, faster, and safer to go from prompt to production.

You’ve probably experienced this: your AI agent in Cursor generates a perfect React app, complete with components and thoughtful styling. Then it stops. ‘To deploy this, you’ll need to create a Netlify account, set up your Git repository…’

The AI magic ends. That’s a poor agent experience. Instead, what if your coding agent could go from generating code to deploying that code in the same conversation?

Agents that ship

Most AI development tools stop at codegen. They write your application, suggest improvements, even explain how deployment works, but they can’t actually deploy your code. You’re left to manually bridge the gap between agent-generated code and production.

The Netlify MCP Server changes this, drastically improving the experience for you and your agents. With it, agents can complete the entire development cycle all within your editor and workflow.

Introducing the official Netlify MCP Server

Today we’re releasing the Netlify MCP Server. Built on Anthropic’s open source Model Context Protocol, our MCP Server gives agents direct access to the Netlify API and CLI. Agents can now create projects, deploy applications, and manage your infrastructure using natural language prompts.

What agents can now do

With the Netlify MCP Server, you can ask your agent to:

Create and manage projects

Create new Netlify projects directly from your editor
Deploy your code with a simple prompt
Manage project settings and configurations

Secure and configure

Set up access controls to protect projects from unwanted access
Create environment variables and secrets for projects
Install or uninstall Netlify extensions

Handle team workflows

Fetch user and team information
Enable and manage form submissions on projects
Complete deployment workflows without leaving your development environment

Complete the development cycle

Generate code, deploy it, and iterate within the same conversation
Move from “here’s your code” to “here’s your live URL”

Getting started

The Netlify MCP Server works with MCP-compatible coding agents like Windsurf, Cursor, Claude, and VSCode Copilot. Setup takes minutes:

Install Node.js 22 or higher
Add the MCP Server to your editor’s configuration
Start using natural language prompts to deploy with Netlify

This is an example of a mcp.json file that adds the Netlify MCP server:

{
  "mcpServers": {
    "netlify-mcp": {
      "command": "npx",
      "args": ["-y", "@netlify/mcp"]
    }
  }
}

Your agent can immediately start creating projects, deploying code, and managing your Netlify infrastructure.

From prompt to code to live URL

This is part of the agent experience that sets Netlify apart when building with AI. While other platforms require manual handoffs between code generation and deployment, Netlify enables agents to complete the full journey from prompt to production.

Tomorrow we’ll address another piece of the prompt-to-production puzzle: where do AI-generated applications store their data?

The official Netlify MCP Server is available now. Connect it to with your favorite agent and give them the power to ship code faster.

推荐订阅源