@jatidevelopments
基本信息
Advanced Puppeteer automation server with enhanced functionality (MCP Server)
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"mcp-puppeteer-advanced": {
"command": "docker",
"args": [
"build",
"-t",
"mcp/puppeteer",
"-f",
"Dockerfile",
"."
]
}
}
}Navigate to any URL in the browser
Capture screenshots of the entire page or specific elements
Click elements on the page
Hover elements on the page
Fill out input fields
Select an element with SELECT tag
Execute JavaScript in the browser console
Extract all images from the page (both `<img>` tags and CSS background images)
Download extracted images to a specified folder
Analyze a DOM element to extract its HTML structure, Markdown representation, and applied styles
Check browser status, list open tabs, and manage tab switching
Analyze the DOM hierarchy of a page and determine parent-child relationships between elements
Switch the browser viewport to simulate different devices (mobile, tablet, desktop)
Navigate back or forward in browser history
Puppeteer is a Model Context Protocol server that provides browser automation capabilities using Puppeteer. It enables LLMs to interact with web pages, take screenshots, extract and download images, and execute JavaScript in a real browser environment.
Configure the server in your MCP client using either Docker (docker run -i --rm --init -e DOCKER_CONTAINER=true mcp/puppeteer) or NPX (npx -y @modelcontextprotocol/server-puppeteer). Customize Puppeteer’s browser behavior via the PUPPETEER_LAUNCH_OPTIONS environment variable or by passing launchOptions and allowDangerous to the puppeteer_navigate tool.
<img> tags and CSS backgroundsPuppeteer offers tools for navigation, screenshots, clicking, hovering, form filling, selecting, evaluating JavaScript, extracting/downloading images, analyzing DOM elements and page hierarchy, managing browser tabs, switching viewports, and navigating history.
Set the `PUPPETEER_LAUNCH_OPTIONS` environment variable as a JSON-encoded string, or pass `launchOptions` and `allowDangerous` parameters directly to the `puppeteer_navigate` tool.
Two resource types: `console://logs` (browser console output) and `screenshot://<name>` (PNG images of captured screenshots).
Yes. Use the Docker configuration shown in the README. The Docker version runs headless Chromium, while the NPX version opens a browser window.
Yes, but they are disabled by default. Set `allowDangerous` to `true` (either in the environment via `ALLOW_DANGEROUS` or in the tool call) to use options like `--no-sandbox` or `--disable-web-security`.
Let Claude Code, Claude Desktop, OpenAI Codex, Kimi and any MCP-capable AI agent create, launch, check and drive browser profiles in Nullprint, the AI antidetect browser. Install with one command.
Playwright Model Context Protocol Server - Tool to automate Browsers and APIs in Claude Desktop, Cline, Cursor IDE and More 🔌
Monitor browser logs directly from Cursor and other MCP compatible IDEs.
Playwright MCP server
Give your AI agent the live web. One MCP server for Search, Fetch, Browser, and Agent: browse real sites, extract structured data, fill forms, and run multi-step workflows. Search and Fetch are free.
评论