Easily Start Browser Automation with MCP: Puppeteer Server Setup

**Want to start browser automation using MCP (Model Context Protocol)? Forget confusing setup steps! In this guide, I'll show you step by step how to run an MCP server with Puppeteer support with just a single npx command. And all this on Windows environment! 💻

By reading this article to the end, you'll understand how LLMs (Large Language Models) perform operations through a real browser and be able to easily integrate it into your own development environment.

What is MCP and Why is it Important?#

Model Context Protocol (MCP) is a protocol that facilitates communication between artificial intelligence models and various context providers (code editor, terminal, browser, etc.).

So why is it so important?

• 🤖 Enables LLMs to navigate in browsers.
• 🧠 Provides advanced code completion and documentation capabilities.
• 🔧 Increases productivity with end-to-end integration.

What is Puppeteer?#

Puppeteer is a Node.js library that allows you to programmatically control Chrome/Chromium browser.

With Puppeteer: - You can navigate web pages, - Take screenshots, - Fill out forms, - Crawl page content.

And with MCP, it's possible to delegate this to LLMs!

Problem: Lack of Quick Setup#

Starting browser automation usually requires complex configurations. This can be discouraging for beginners.

Also, the wrong npx command can frustrate you with errors like entrypoint not found. The purpose of this guide is to eliminate this problem with a simple and working solution.

Solution: Starting Puppeteer Server with Single Command ✅#

1. Prerequisites#

Make sure the following software is installed on your system:

• Node.js (at least v18+)
• npx command (comes with Node)

To check, type this in terminal:

node -v
npx -v

2. Starting Server with npx#

The command is very simple. Open terminal in Windows environment and type the following:

npx -y @modelcontextprotocol/server-puppeteer

🚀 This command downloads the required package and starts the browser-supported MCP server. When it works correctly, you'll see output similar to this in terminal:

[MCP] Puppeteer server listening on port 4455...

Browser also opens automatically (not headless).

3. Configuration with YAML (Optional)#

For advanced usage, you can add this server to mcpServers blocks:

mcpServers:
  - name: puppeteer
    command: cmd
    args:
      - /c
      - npx
      - -y
      - "@modelcontextprotocol/server-puppeteer"

I'm using it in Continue plugin as follows, together with ollama

name: Local Assistant
version: 1.0.0
schema: v1
models:
  - name: deekseek-r1
    provider: ollama
    model: deepseek-r1
  - name: Qwen 2.5 Coder 7b
    provider: ollama
    model: qwen2.5-coder:7b
    roles:
      - autocomplete
context:
  - provider: code
  - provider: docs
  - provider: diff
  - provider: terminal
  - provider: problems
  - provider: folder
  - provider: codebase
docs:
  - name: Pelican
    faviconUrl: ""
    startUrl: https://docs.getpelican.com/en/latest/quickstart.html

mcpServers:
  - name: browser-tools
    command: cmd
    args:
      - /c
      - npx
      - -y
      - "@agentdeskai/browser-tools-mcp@1.2.0"
  - name: puppeteer
    command: cmd
    args:
      - /c
      - npx
      - -y
      - "@modelcontextprotocol/server-puppeteer"

This way config-based startups are possible. 🔧

Common Errors and Solutions 🧯#

mcp-server-puppeteer is not recognized...#

This error usually occurs when npx package is called incorrectly. Only solution:

✅ Run @modelcontextprotocol/server-puppeteer package directly with npx.

Real Life Usage Scenarios#

• 📄 Retrieving dynamic content from documents with browser
• 🛠️ Making web application tests LLM-based
• 💬 Chatbots retrieving information from live pages

With MCP, all these operations can be done without writing any extra code!

Conclusion and Call to Action 🎯#

Starting MCP with Puppeteer server has never been this easy.

You can set up a powerful browser-supported LLM environment with a single command. By applying this article, you can take a step ahead in your code-based projects.

📣 If this guide helped you, don't forget to leave a comment and follow the blog! I'll explain more LLM integrations in new articles.

If you're ready, now open your terminal and run the command:

npx -y @modelcontextprotocol/server-puppeteer

Happy coding! 🚀