Home / MCP / Scrape Copilot MCP Server

Scrape Copilot MCP Server

Provides browser automation and web interaction via Puppeteer for LLMs to navigate pages, take screenshots, and run JavaScript.

javascript
Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
    "mcpServers": {
        "puppeteer": {
            "command": "docker",
            "args": [
                "run",
                "-i",
                "--rm",
                "--init",
                "-e",
                "DOCKER_CONTAINER=true",
                "mcp/puppeteer"
            ],
            "env": {
                "DOCKER_CONTAINER": "true"
            }
        }
    }
}

You can run a Puppeteer-based MCP Server that exposes browser automation capabilities to enable LLMs to navigate web pages, take screenshots, click elements, fill forms, and run JavaScript in a real browser environment. This server exposes a set of practical tools you can invoke from your MCP Client to automate web tasks safely and reproducibly.

How to use

To use this server, connect your MCP Client to either the Docker-based runtime or the NPX-based runtime. You will access a set of browser-automation tools that let you navigate to URLs, capture screenshots, interact with page elements, and execute JavaScript in the browser. The two main tools you leverage are for navigation and interaction, plus utilities to inspect results via console logs and screenshots.

Core tools you will use include navigational, interaction, and evaluation capabilities. The common pattern is to issue a command to the corresponding tool with the required inputs, then inspect the results through browser console output or captured screenshots.

How to install

Prerequisites: you need Docker or Node with NPX installed, depending on your preferred runtime.

Install and run using Docker (headless Chromium): copy the following command sequence and execute it in your terminal.

How to install

Build the Docker image for the Puppeteer MCP server if you prefer a containerized runtime.

docker build -t mcp/puppeteer -f src/puppeteer/Dockerfile .

Available tools

puppeteer_navigate

Navigate to a URL in the browser. Input: url (string)

puppeteer_screenshot

Capture screenshots of the entire page or a specific element. Inputs: name (string, required), selector (string, optional), width (number, optional, default 800), height (number, optional, default 600)

puppeteer_click

Click elements on the page. Input: selector (string)

puppeteer_hover

Hover elements on the page. Input: selector (string)

puppeteer_fill

Fill out input fields. Inputs: selector (string), value (string)

puppeteer_select

Select an element with SELECT tag. Inputs: selector (string), value (string)

puppeteer_evaluate

Execute JavaScript in the browser console. Input: script (string)