browser-automation--web-scraping.md

🌐 Browser Automation & Web Scraping

Servers using tools for browser control, automation, and extracting content from websites.

• jiawei686/jev-ultrafast-mcp: Hands a whole browser task off in one call instead of a turn per click — a decision model drives the loop server-side over ref-based element tables, assertions are code-checked rather than model-judged, and a recorded path replays at zero model cost. Speaks CDP to a Chrome you already have: no Playwright, no screenshots. Python, MIT licensed, on PyPI.

• usestring/string-ai-mcp: Search the web, fetch any URL and map a site — clean Markdown, past anti-bot blocks. Proxy rotation, anti-bot handling, CAPTCHA solving and JavaScript rendering happen server-side, so the agent gets the page instead of a block screen. Tools: web_access_search, web_access_fetch, web_access_sitemap. Install: npx -y @usestring/mcp with STRING_AI_API_KEY; hosted Streamable HTTP endpoint at https://mcp.usestring.ai/v1/mcp (OAuth). MIT. Registry id ai.usestring/web-access.

• Hronaut: Source-available Electron/Chromium desktop browser workspace for local MCP clients, with named persistent tabs and storage and visible human handoff for sign-in, 2FA, CAPTCHA, and consequential writes. Windows, macOS, and Linux downloads: setup and downloads; Hronaut Subscription and Trial License, with a 10-day trial followed by $4/month or $24/year per named user.

• leek-emperor/Personae: MCP server for controlling a specific already-open, account-isolated Personae browser identity instead of launching a separate browser. Supports navigation, accessibility snapshots, interaction, screenshots, JavaScript evaluation, and OAuth popup targets. stdio transport; no API key or public listener. Install: npx -y @leek-emperor/personae-mcp after installing and launching Personae; MIT.

• rasterly-dev/rasterly-mcp: MCP server for rasterly.dev that gives agents five web-rendering tools: read a JavaScript-rendered page as clean Markdown, capture a screenshot, render a PDF, extract structured JSON, or crawl same-origin pages into Markdown. Install: npx -y rasterly-mcp; requires RASTERLY_API_KEY; MIT.

• VibeTechnologies/vibe-mcp: Drives your real, already-logged-in Chrome from any MCP client. The agent can run on a different machine and reaches the browser over an outbound WebSocket relay, so no inbound port has to be open; uses your real Chrome profile with no per-connection approval dialog, and several agents can share one browser. 28 tools (navigate, click, fill, snapshot, screenshot, network, evaluate). stdio transport, no auth for local use. Install: npx -y @vibebrowser/mcp@latest. Apache-2.0.

• Satyajeet-04/parley: Lightweight CDP browser-automation framework with 14 MCP tools — DOM reading, cookie extraction, JS execution, and AI chat bridging across ChatGPT/Gemini/Claude tabs. Zero heavy deps, text-only (no screenshots).

• unitedideas/actablesite-mcp: Read-only website audit MCP server for AI-crawler policy checks and full-report handoff. Supports Streamable HTTP and a dependency-free stdio bridge with no authentication; exposes 3 bounded tools and cannot modify sites. MIT.

• rog0x/mcp-web-tools: Web MCP tools for scraping, search, and monitoring workflows. Install from the @rog0x npm package family. MIT.

• rog0x/mcp-seo-tools: SEO MCP tools for metadata, keywords, page speed, and search visibility checks. Install from the @rog0x npm package family. MIT.

• Crawlio-app/crawlio-browser-agent: AI-powered website crawling MCP server with 38 tools for crawl control, browser enrichment capture, export (WARC/ZIP/single-HTML), observation timeline, and evidence-backed findings. Install: npx crawlio-mcp.

• Komos MCP: Hosted MCP server for creating, running, and monitoring browser workflow automations with integrations, files, web search, and email tools.

• Quality-Max/qmax-mcp: Gives coding agents independent QA evidence before declaring a web change done — scans a URL for console errors, broken links, accessibility issues, Core Web Vitals/performance, SEO, security headers, cookies/trackers, mixed content, and page weight; inspects page structure for locator candidates; generates a deterministic Playwright repro; and runs one Playwright test behind a human-approval gate. Install: npx -y @qualitymax/qmax-mcp; requires Node.js >=22.13.0; MIT.

• giannisalinetti/python-mcp-server: Facilitates Python code execution for web scraping tasks using an LLM, leveraging Podman for container management.

• mhazarabad/browser-use-mcp: Automates browser tasks using the Browser Use API, offering task management and monitoring capabilities.

• luminati-io/web-scraping-with-mcp: Connects Anthropic LLMs to external web scraping tools using Bright Data integration for real-world data extraction.

• magicp-mhl/newsnow: Facilitates elegant real-time news reading with adaptive scraping and MCP server support for seamless integration.

• regenrek/deepwiki-mcp: Fetches and converts Deepwiki content into Markdown for LLM consumption, ensuring domain safety and efficient crawling.

• yan5236/bing-cn-mcp-server: A Chinese Bing search tool leveraging MCP for seamless integration with AI tools like Claude, enabling direct Bing searches and webpage content retrieval without an API key.

• w1561778301/mcp-playwright-test: Automates Playwright tests by generating and executing UI and API test cases with detailed reporting.

• xinlei413/DOC-Server-MCP: Fetches and searches third-party package documentation using versatile scraping and intelligent processing with hybrid search capabilities.

• eadm/grain-mcp-server: Facilitates integration with Grain for meeting recording and transcription through browser automation, bypassing enterprise API restrictions.

• jobsonlook/xhs-mcp: Facilitates interaction with Xiaohongshu by reverse-engineering JavaScript to access and manage notes and comments without Playwright.

• myzxlin/redbook-mcp: Facilitates automated publishing of text and video notes on Xiaohongshu using Playwright for browser automation.

• larryhudson/mcp-server-example-image-block: Provides a random image retrieval service using the Lorem Picsum API, demonstrating MCP server capabilities with image blocks.

• Elvis720/mcp-agent: Facilitates browser automation for LLMs using Playwright's accessibility tree, enabling structured web interactions without visual models.

• xinlei413/MCP-DOC-Server-OpenRouter: Facilitates efficient scraping, processing, and searching of third-party package documentation with advanced hybrid search capabilities.

• thomasvan/mcp-brave-search: Facilitates AI assistant integration with Brave Search through a Model Context Protocol server.

• manchittlab/TheCrawler: Web crawler and extractor MCP server with markdown output, PDF/DOCX parsing, structured errors, and validated extraction contract diagnostics for agents.

• oddlyspaced/ultimate-android-mcp: Empower LLMs to autonomously manage and interact with Android devices through a comprehensive MCP server.

• ronantakizawa/a11ymcp: Empowers LLMs to conduct comprehensive web accessibility testing using Deque Axe-core API and Puppeteer.

• hamibot/hamibot-mcp-server: Facilitates interaction with the Hamibot API, offering tools for device and script management, including custom JavaScript execution on specified devices.

• YantaoMou/mcp_droid: Facilitates direct control and operation of Android devices by large language models through a standardized interface.

• LuckyXYJ/mcp_ios_project: An MCP server configured for iOS projects, utilizing Python scripts for server execution.

• ying-dao/yingdao_mcp_server: Empowers YingDao AI Power and other MCP Hosts to execute RPA tasks, supporting both SSE and Stdio server modes.

• catinair/aipower-rpa-mcp-server: Facilitates seamless interaction between YindDao AI Power and RPA tools, enabling AI-driven automation workflows.

• bneil/mcp-go-colly: A sophisticated web crawling framework integrating MCP with Colly for flexible and extensible web content extraction in LLM applications.

• ertugrul59/tradingview-chart-mcp: Fetches TradingView chart images using ticker and interval inputs for seamless integration with MCP clients.

• sungithubid/mcp-url2markdown: Transforms web page content from a given URL into clean, formatted markdown using MCP server capabilities.

• NexusX-MCP/data-mcp-server: Facilitates web scraping, data extraction, and browser automation with integration for agents like OpenAI's CUA and Anthropic's Claude.

• korwabs/playwright-trace-mcp: Enhances browser automation with Playwright by adding trace viewer and video recording capabilities, enabling LLMs to interact with web pages through structured data.

• jarondonp/airbnb: Facilitates searching and retrieving detailed information on Airbnb listings without requiring an API key.

• code-creations-io/build-langchain-web-browsing-agent-with-mcp: Leverage LangChain and MCP servers to build web browsing agents for Airbnb and Google Chrome using Python and Playwright.

• moonbirdai/puppeteer-plus-martech-mcp: Enhances Puppeteer for digital marketing and SEO analysis, enabling AI assistants to audit and detect marketing technologies on websites.

• zueai/webdev-mcp: Provides web development tools for taking and listing screenshots via an MCP server.

• TheQtCompanyRnD/squish-mcp: Facilitates AI-driven execution and analysis of Squish test scripts for automated GUI testing.

• johannesPettersson80/codesys-mcp-toolkit: Facilitates seamless automation of project management and code editing tasks in CODESYS environments through MCP client interactions.

• korwabs/playwright-record-mcp: Enables LLMs to automate browser interactions with video recording using Playwright.

• crying-pizza/mcp-naver-tennis-reservation: Facilitates automated checking of available time slots at Yangjae Tennis Court via Naver Reservation using Claude integration.

• AppiumTestDistribution/mcp-webdriveragent: Facilitates the building and signing of WebDriverAgent for iOS, streamlining its preparation for Appium integration.

• hao-cyber/phone-mcp-by-trae: Control Android phones via ADB commands using a Trae-based MCP plugin.

• hoainho/podium-mcp: Unified mobile E2E testing MCP server wrapping Maestro. Execute Maestro flows on iOS/Android simulators and devices, manage test suites, and integrate with CI/CD pipelines.

• chenningling/Redbook-Search-Comment-MCP2.0: A Playwright-based tool for automating Xiaohongshu searches and AI-generated comments, seamlessly integrating with MCP Clients like Claude for Desktop.

• ratiofu/mcp-puppeteer: Facilitates Chrome control through Puppeteer using an MCP server setup.

• DannyAtVodooTH/ai_browse: Empower Claude Desktop with web browsing and interaction capabilities through a lightweight MCP server.

• rascal-3/MCP-servers: Integrates the Sonar API to enable Claude with real-time, web-wide research capabilities.

• PyneSys/claude_autoapprove: A CLI tool for automatically approving tool requests in the Claude Desktop App using remote debugging.

• ciaraadkins/puppeteer-plus-mcp: Enhances Puppeteer with advanced digital marketing and SEO analysis capabilities, enabling comprehensive audits of website marketing technologies and SEO practices.

• chenningling/RedBook-Search-Comment-MCP: Automates login, search, and intelligent commenting on Xiaohongshu using Playwright, configurable as an MCP Server for seamless client integration.

• cdecl/oo.ai-mcp: Facilitates content search on oo.ai through an MCP server setup, leveraging fastmcp and Playwright.

• KiyotakaMatsushita/mcp-youtube: MCP-YouTube leverages yt-dlp to enable seamless video, audio, and subtitle downloads from YouTube via a single JSON call.

• pietrozullo/browser-use-mcp: FastMCP server enabling browser automation via natural language commands, allowing Language Models to perform web-based tasks through a simple API.

• yhanawa/mcp_server: Facilitates API documentation access through an MCP server, utilizing scripts for web crawling and JSON/YAML data handling.

• datptrunsystem/PlaywrightMCP: Enables LLMs to perform browser automation tasks using Playwright's accessibility tree for structured web interactions.

• rainyheart/jcrawl4ai-mcp-server: A Java-based MCP server leveraging Spring Boot to facilitate web crawling through the Crawl4ai API.

• Wh1isper/url-instax: Facilitates LLMs in capturing webpage screenshots via a simple MCP server setup.

• Redtri-git/Ecomm-MCP: Facilitates product search through a minimal MCP server, integrating seamlessly with Claude for querying product listings via FastMCP.

• rodrigopv/nextr4y: A Golang-based tool for remote scanning and integration of Next.js applications, revealing build IDs, routes, and assets while evading bot detection.

• keygray/duckduckgo-server: Provides web search capabilities through DuckDuckGo with advanced rate limiting, content fetching, and LLM-friendly output.

• hunter-arton/google_search_mcp_server: Enables real-time web and image search capabilities for AI assistants using Google's Custom Search API.

• SarthakMishra/site-cloner: Facilitates website cloning by providing tools to fetch, analyze, and download website assets for LLMs.

• suvraadeep/Airbnb-mcp: Integrates Airbnb MCP servers with LangChain and Groq through a browser interface for seamless interaction.

• JonathanHsuHH/web-crawler-mcp-server: Facilitates web crawling and content extraction for AI assistants using Puppeteer to bypass anti-bot measures.

• ikeyu0806/my-mcp-server: A personal MCP server for executing news collection and web scraping tasks with customizable settings for Claude Desktop and Cursor.

• GitDzreal93/uiautomator2-mcp: Facilitates Android device automation through UIAutomator2, enabling control over device management, UI operations, and application management.

• mugoosse/sitemap-mcp-server: Fetch, parse, and visualize website sitemaps to uncover hidden pages and analyze site structures efficiently.

• panxingfeng/mcp_server_searXNG: Facilitates privacy-friendly web searches using SearXNG with customizable search engines and categories.

• flyiu/mobiler-use: A Java-based mobile device control service leveraging Spring Boot and Appium, enabling REST API and MCP protocol interactions for Android devices.

• dragons96/mcp-undetected-chromedriver: A robust MCP service leveraging undetected-chromedriver to automate Chrome browser tasks while bypassing anti-bot detection.

• l0kifs/browser-mcp: A robust MCP server for browser automation using Playwright, enabling scriptable and reliable DOM interaction and automation.

• Hisma/openrouter-search: Leverages the OpenRouter API to enable web search capabilities through an MCP server using the google/gemini-2.5-pro-preview-03-25:online model.

• modularflow/mcp-ui-explorer: Facilitates interaction with UI elements on your screen, enabling exploration, visualization, and automation of UI tasks.

• cthuaung/mcp-use-cli: An interactive command-line chat application leveraging MCP for AI-driven web browsing and conversation memory.

• Godzilla675/doc-context-provider-mcp: Crawls and summarizes online documentation using the Gemini API, providing context for MCP clients.

• markbustamante77/mcp: Enables LLMs to perform browser automation tasks using Playwright's accessibility tree for structured web interactions.

• swasthik-007/MCP_trial: Integrates Playwright, Airbnb, and DuckDuckGo Search agents for automation, real estate search, and web queries within the Cursor IDE.

• manoj9788/mcp-axe: Automated accessibility scanning plugin using axe-core, compatible with MCP, supporting Selenium and Playwright engines.

• sinco-lab/mcp-config-manager: A browser extension that simplifies MCP server configuration management with an intuitive interface and cross-platform support.

• browserstack/mcp-server: BrowserStack MCP Server integrates with AI tools to facilitate comprehensive manual and automated testing across real devices and browsers.

• JonaFly/RednoteMCP: Automates Xiaohongshu login, keyword search, note retrieval, and intelligent commenting using Playwright.

• en30/mcp-cosense-server: Facilitates LLM interactions with Cosense through Puppeteer-driven browser automation for page creation, modification, and search.

• clareliguori/wordhippo-mcp-server: A specialized MCP server that retrieves thesaurus data from WordHippo.

• MDBs123/playwright-mcp: Enables LLMs to perform browser automation tasks using Playwright's accessibility tree for efficient web interactions.

• webscraping-ai/webscraping-ai-mcp-server: Integrates with WebScraping.AI to provide advanced web data extraction capabilities using the Model Context Protocol.

• luminati-io/brightdata-mcp: Bright Data's MCP server facilitates seamless public web data access for decision-making by various MCP clients.

• itsShashankSrivastava/Linkedin-Scrap-MCP-Server: Fetches real-time LinkedIn profile data using an MCP server integrated with the Fresh LinkedIn Profile Data API.

• dlwjdtn535/mcp-chrome-server: Facilitates browser automation through a Chrome server using the MCP model, enabling seamless web interactions and credential management.

• kentaro84207/mcp-sitefetch: Fetches entire websites and integrates their content into LLM contexts for enhanced analysis and interaction.

• snd-primary/web-scraper: A Python-based MCP server offering web scraping capabilities from MDN Web Docs, seamlessly integrating with Claude Desktop.

• stickerdaniel/linkedin-mcp-server: Facilitates LinkedIn data scraping and job searches through AI assistants like Claude.

• lololosys/Unblocker-MCP: Access blocked websites by bypassing anti-bot systems using NetNut's proxy service with configurable geo-location and HTML rendering options.

• dlwjdtn535/mcp-chrome-integration: Enables AI models to automate and control Chrome browser tasks through a protocol for seamless web interactions.

• qoxogus/spring-ai-chrome-mcp-server: Facilitates interaction with Chrome browser through a Spring-based MCP server, enabling automated tasks like searching, navigating, and summarizing content.

• xingzhicn/playwright-mcp-docker: Effortlessly set up and manage a Playwright MCP server using Docker Compose for seamless integration with clients like Cline or Cursor.

• Streamline-TS/mcp-server-flyder: Facilitates Flyder workflow integration through a Model Context Protocol server.

• axlwolf/brave-mcp: Integrates the Brave Search API to provide comprehensive web and local search capabilities with smart fallbacks and flexible filtering.

• 302ai/302_browser_use_mcp: Facilitates browser automation through MCP integration, compatible with Claude Desktop.

• vishwajeetdabholkar/eget_mcp: Integrates eGet web scraper with Claude for Desktop, enabling seamless web content scraping and analysis through a local API.

• kazuph/mcp-qdrant-docs: Scrapes website content, indexes it into Qdrant, and enables natural language queries about the indexed content.

• DevEnterpriseSoftware/scrapi-mcp: Leverage ScrAPI for seamless web scraping with advanced bot detection and geolocation bypass capabilities.

• dhruv01234/Mcp-playwright: Facilitates browser automation using Playwright with MCP server capabilities, supporting both snapshot and vision-based interactions.

• vikaskumarprofessional92/playwright-mcp-project: Demonstrates the integration of a Playwright MCP server with Jenkins for automated browser testing.

• hdcola/mcp-server-playwright: Facilitates browser control via AI using Playwright through MCP integration.

• djyde/browser-mcp: Enhance browser interaction with a versatile MCP server and extension, enabling tasks like page summarization, style modification, and history search.

• madtank/nova-act-mcp: Facilitates multi-step browser automation workflows using the Amazon Nova Act SDK, enabling AI assistants to control web browsers with transparent agent reasoning.

• hieu-blackbox/mcp-puppeteer: Enables LLMs to automate web interactions using Puppeteer, including navigation, screenshots, and JavaScript execution.

• hieu-blackbox/mcp-browserbase-stagehand: Empowers LLMs with AI-driven web automation through Stagehand, enabling interaction, data extraction, and action observation in a real browser environment.

• Shubhwithai/linkedin-scraper-mcp: Efficiently extract LinkedIn profile data using an MCP server with asynchronous operations and streamlined JSON output.

• pinkpixel-dev/prysm-mcp-server: Prysm MCP Server empowers AI assistants to perform high-accuracy web scraping with multiple extraction modes and flexible output formats.

• vpsone/browser-control-mcp: Empowers LLM clients to control local Firefox browsers by managing tabs, browsing history, and web content through an MCP server and browser extension.

• infinitimeless/podcrawler-mcp: PodCrawlerMCP facilitates podcast discovery by crawling the web for RSS feeds and integrating seamlessly with AI assistants.

• Yazington/docs-crawler: Enables crawling and searching of documentation websites using vector embeddings for efficient content retrieval.

• themindmod/selenium-mcp-server: Facilitates AI-driven browser automation and web interaction using Selenium WebDriver through the Model Context Protocol.

• loo-y/mcp-smart-crawler: MCP Smart Crawler leverages Playwright to automate web crawling, metadata extraction, and resource downloading from Xiaohongshu posts.

• canadaduane/browse-together-mcp: Facilitates co-browsing with AI by integrating a Playwright-controlled browser and MCP server for seamless browser automation and interaction.

• hetaoBackend/mcp-pyautogui-server: Automate GUI testing and control across platforms using PyAutoGUI with this MCP server.

• coucya/mcp-server-requests: Facilitates LLM-driven web content retrieval and processing with comprehensive HTTP request capabilities.

• Yazington/docs-crawler-mcp: Crawls documentation websites, converts them to Markdown, and stores embeddings in Qdrant for intelligent search.

• ndthanhdev/mcp-browser-kit: Facilitates interaction with manifest v2 compatible browsers through an MCP server, supporting integration with Brave and Firefox.

• puremd/puremd-mcp: Enhance MCP clients with web unblocking and search capabilities using pure.md's markdown delivery network.

• matrupriya2048/linkedin-profile-scraper: Fetch LinkedIn profile data asynchronously using the Fresh LinkedIn Profile Data API, with environment-based configuration and MCP framework integration.

• runablehq/mobile-mcp: Facilitates mobile automation by enabling LLMs to interact with Android devices through structured UI dumps.

• iota31/BrowserToolsMCP: Facilitates browser interactions for debugging and testing web applications through Claude, supporting multiple browsers and offering tools like navigation, screenshots, and performance tests.

• scrapeless-ai/scrapeless-mcp-server: Integrate real-time Google SERP results into LLM applications for dynamic context retrieval in AI workflows and chatbots.

• n11west11/puppeteer-with-authentication: Enables LLMs to automate browser tasks using Puppeteer, including navigation, interaction, and JavaScript execution.

• egorov-ma/mcp-selenium-test: Automate browser testing using mcp-selenium, a Model Context Protocol implementation for Selenium WebDriver, enabling cross-platform browser control via API without local driver installations.

• ShreeraghavM/executeautomation1: Enables LLMs to perform browser automation tasks using Playwright, including web scraping and JavaScript execution.

• burnworks/burnworks-mcp-server-fetch: Enables LLMs to fetch and process web content using CSS selectors, converting HTML to markdown for efficient data extraction.

• AdsPower/local-api-mcp-python: Facilitates interaction with AdsPower's LocalAPI for browser management via LLMs.

• didibao/puppeteer-mcp: Enables LLMs to automate browser tasks using Puppeteer, including navigation, interaction, and JavaScript execution.

• Aosamesan/kotlin-mcp-playwright-test: A Kotlin-based MCP server leveraging Playwright for browser automation testing.

• bluewd111/e2e-by-mcp: Generate end-to-end tests for web applications using Playwright MCP.

• jasondsmith72/Browser-use-claude-mcp: Enables AI models like Claude and Gemini 2.5 to perform browser automation and web interactions through natural language commands.

• kitfactory/PyMCPAutoGUI: PyMCPAutoGUI enables AI agents to automate GUI tasks by interacting with desktop applications through MCP, enhancing productivity and testing capabilities.

• iuill/playwright-mcp-docker: Facilitates the deployment of a Playwright MCP server using Docker Compose for seamless client integration.

• DevMorph77/mcp-playwright: Integrates Playwright with MCP for seamless web automation and testing through natural language or programmatic interaction.

• ruifigueira/playwright-mcp: Enables LLMs to automate browser interactions using Playwright's accessibility tree for efficient and structured web page manipulation.

• ziux/playwright-plus-python-mcp: Enhances browser automation capabilities using Playwright with a focus on note management and advanced page interactions.

• mobile-next/mobile-mcp: Facilitates scalable mobile automation by enabling interaction with native iOS and Android applications through a platform-agnostic interface.

• hathibelagal-dev/mcp-pyautogui: Facilitates desktop automation by enabling AI systems to control mouse, keyboard, and screen interactions using PyAutoGUI.

• jamesarnold0816/firecrwal: Integrates with Firecrawl for advanced web scraping, crawling, and content extraction with support for cloud and self-hosted instances.

• eyalzh/browser-control-mcp: Facilitates LLM-driven browser control through an MCP server and Firefox extension, enabling tasks like tab management and web content interaction.

• haviethung3004/MCP_Network_automator: An AI-assisted tool for automating Cisco network device management through natural language commands, integrated with Claude Desktop.

• Ichigo3766/crawl4ai-mcp: Provides web crawling capabilities with markdown output and citations for LLM integration.

• toliuweijing/mcp-browser-automation: A TypeScript-based MCP server that implements a simple notes system with tools for creating and summarizing text notes.

• xkiranj/playwright-universal-mcp: A versatile Playwright MCP server enabling browser automation across multiple environments, including containerized setups, with support for various browsers and integration with AI assistants.

• Selenium39/mcp-server-weibo: A TypeScript-based server for scraping Weibo user information, feeds, and search functionality using the Model Context Protocol.

• aptcats/mcp-server-playwright: A browser automation service leveraging Playwright for seamless integration with Claude Desktop.

• tanob/mcp-desktop-automation: Empower LLMs with desktop automation through mouse and keyboard control, and screenshot capabilities using RobotJS.

• Abhi5h3k/MCP-URL2SNAP: A lightweight MCP server that captures webpage screenshots and returns access URLs, enhancing AI applications with visual capture capabilities.

• nathanonn/mcp-url-fetcher: Fetches and converts web content from any URL into various formats, enhancing LLMs' ability to access and analyze web data.

• microsoft/playwright-mcp: Facilitates browser automation for LLMs using Playwright's accessibility snapshots, enabling efficient web interactions without visual models.

• browsercat/browsercat-mcp-server: Enables LLMs to automate web interactions and execute JavaScript using BrowserCat's cloud service without local browser installation.

• lxe/chrome-mcp: Enables precise control over a Chrome browser instance via the Chrome DevTools Protocol, facilitating seamless integration with LLM agents.

• thirdstrandstudio/mcp-xpath: Facilitates the execution of XPath queries on XML and HTML content, supporting both local and URL-based data sources.

• gpaul-faldin/MCP_puppeteer_extra: Enhances browser automation with Puppeteer-Extra and Stealth Plugin for LLMs to interact with web pages undetected.

• 9pros/MCP-Browser-Inspector: Facilitates web element inspection and responsive testing with interactive messaging and modification capabilities.

• jae-jae/fetcher-mcp: Fetcher MCP leverages Playwright to efficiently extract and process dynamic web content, offering intelligent content extraction and flexible output formats.

• Scrapybara/scrapybara-mcp: Facilitates interaction with virtual Ubuntu desktops for web browsing and code execution via MCP clients.

• xxxbrian/mcp-rquest: Enables LLMs to perform realistic browser-like HTTP requests with advanced fingerprinting to bypass anti-bot measures and convert documents to Markdown for enhanced processing.

• MaartenSmeets/mcp-server-fetch: Fetches and processes web content using browser automation and OCR, enabling LLMs to access complex web pages.

• cploujoux/mcp-puppeteer: Enables LLMs to automate browser tasks using Puppeteer, including navigation, interaction, and JavaScript execution.

• yokingma/one-search-mcp: OneSearch MCP Server integrates multiple search engines and web scrapers for comprehensive web search, scraping, and extraction capabilities.

• janspoerer/mcp_browser_use: Empowers AI agents to perform browser automation and web scraping using MCP and Selenium.

• PatrickAbainza/playwright-plus-python-mcp_my_version: A Playwright-based MCP server offering tools for web navigation, interaction, and content retrieval.

• m-mcp/screenshot-server: A Python-based MCP server offering screenshot capture and image processing capabilities via the MCP protocol.

• mshk/mcp-rss-crawler: Facilitates seamless integration of RSS feeds with LLMs through the Message Chain Protocol.

• win4r/browser-use-MCP-Server: A TypeScript-based MCP server implementing a simple notes system with resources, tools, and prompts for note creation and summarization.

• dazeb/wikipedia-mcp-image-crawler: Facilitates image search and metadata retrieval from Wikipedia Commons, ensuring compliance with Creative Commons licenses for seamless integration into projects.

• wukan1986/mcp_query_table: A Playwright-based financial web table crawler supporting MCP, enabling seamless data extraction from multiple financial websites.

• cyberchitta/scrapling-fetch-mcp: Enables AI assistants to access text content from bot-protected websites using Scrapling, optimized for low-volume retrieval of documentation and reference materials.

• misanthropic-ai/playwrite-mcp: Enables LLM-powered clients to automate browser tasks using Playwright.

• mario-andreschak/mcp-puppeteer-browser: A streamlined MCP server implementation using Puppeteer for browser automation, designed for easy installation and showcasing with FLUJO.

• valyc0/mcp-browser-log: Facilitates browser logging and monitoring through MCP commands, enabling users to list, connect, and retrieve console logs from browser targets.

• kbyk004/playwright-lighthouse-mcp: Analyzes website performance using Playwright and Lighthouse, enabling LLMs to perform detailed performance analysis via MCP.

• snehsuresh/Crawl-In-Chat: A modular system for web crawling and AI-powered chat interfaces, integrating XML sitemap ingestion, vector embeddings, and multiple frontend options including MCP.

• AnuragRai017/Web-Scraper-server: A TypeScript-based MCP server implementing a notes system with resources, tools, and prompts for note management and summarization.

• laurentvv/crawl4ai-mcp: A web crawling tool that seamlessly integrates with AI assistants using MCP to extract and structure website content.

• dennisgl/mcp-playwright-scraper: Effortlessly scrape web content and convert it to Markdown using a Playwright-powered MCP server.

• afshawnlotfi/mcp-configurable-puppeteer: A configurable MCP server leveraging Puppeteer for browser automation, enabling LLMs to interact with web pages and execute JavaScript with customizable options.

• lmcc-dev/mult-fetch-mcp-server: A versatile tool for fetching web content in multiple formats with intelligent proxy detection and bilingual support.

• jroakes/CrawlnChat: A versatile system for web crawling and AI-powered chat interfaces, integrating XML sitemap ingestion, vector embeddings, and multiple frontend options including MCP.

• takezo5096/browser_use_mcp: Facilitates browser control via MCP commands, integrating with LLMs for seamless automation.

• maoxiaoke/mcp-copy-web-ui: Facilitates the download and analysis of complete website content for UI/UX inspiration, making it accessible for AI analysis.

• co-browser/browser-use-mcp-server: Facilitates web browsing through an MCP server with SSE transport, enabling integration with various LLM providers and supporting browser task automation.

• algonacci/mcp-tavily-extract: Facilitates web page extraction for clients through a configurable MCP server.

• ritvij14/crawl4ai-mcp: Integrates Crawl4AI with Cursor AI to provide web scraping and crawling capabilities for LLMs in Cursor Composer's agent mode.

• Ai-Quill/scraperis-mcp: Facilitates AI assistants in web scraping through MCP protocol integration with Scraper.is, offering real-time progress updates and multiple content formats.

• robertheadley/chrome-debug-mcp: Facilitates debugging and automation of Chrome using Playwright with Greasemonkey API support.

• kilicmu/mcp-web-content-pick: Extracts structured content from web pages using customizable selectors and recursive crawling, enhancing web data retrieval capabilities.

• ScrapeGraphAI/scrapegraph-mcp: Enables language models to perform advanced AI-powered web scraping with ScapeGraph API integration.

• web-agent-master/google-search: A Playwright-based tool that enables real-time Google search capabilities for AI assistants by bypassing anti-scraping mechanisms and integrating with MCP servers.

• kevinwatt/playwright-fetch-mcp: Fetch web content in JSON, plain text, or Markdown using Playwright, with support for modern websites and dynamic content.

• bkataru-workshop/mcp-duckduckresearch: Enhances DuckDuckGo search with web content extraction and screenshot capabilities, integrating seamlessly with Roo Code for automated information retrieval.

• amotivv/cloudflare-browser-rendering-mcp: Facilitates web content fetching and processing using Cloudflare Browser Rendering for integration with LLMs in Claude and Cline environments.

• NON906/omniparser-autogui-mcp: Leverage OmniParser to automate GUI operations on Windows through an MCP server.

• Scrapezy/mcp: Enables AI models to extract structured data from websites using the Model Context Protocol.

• bananabit-dev/mcp: Facilitates image generation and web scraping for Windsurf using the Flux Pro model and ScrapeGraph API.

• marcopesani/mcp-server-serper: A TypeScript-based server offering web search and webpage scraping capabilities via the Serper API, integrated with Claude Desktop for enhanced content extraction.

• screenshotone/mcp: Facilitates website screenshot rendering via the ScreenshotOne API using the MCP server protocol.

• stefanskiasan/browser-automation-server: A TypeScript-based MCP server implementing a notes system with resources, tools, and prompts for note management and summarization.

• samcfinan/linear-mcp-server: Enhances project management capabilities for Linear API users through a comprehensive MCP server, facilitating initiatives, projects, and issue management with advanced relationship handling.

• rotexhawk/Mcp-Puppeteer-Server: Enables LLMs to automate browser tasks using Puppeteer, including navigation, interaction, and JavaScript execution.

• DeanXu2357/mcp-web-integration: Facilitates web integration with tools like SearxNG and Crawl4AI for enhanced search and content extraction capabilities.

• angiejones/mcp-selenium: Facilitates browser automation using Selenium WebDriver through standardized MCP clients.

• nguyenvanduocit/fetch-kit: Fetch Kit enables AI models to interact with web services using the Model Context Protocol.

• lars-hagen/mcp-playwright-cdp: Enables LLMs to automate browser tasks and interact with web pages using Playwright and Chrome DevTools Protocol.

• merajmehrabi/puppeteer-mcp-server: Enables browser automation through Puppeteer, facilitating interaction with both new and existing Chrome instances.

• Msparihar/mcp-server-firecrawl: Firecrawl MCP Server offers advanced web scraping, content searching, and site crawling capabilities with robust data extraction features.

• qoli/firefox-headless-server: A TypeScript-based server for managing headless Firefox browser sessions via MCP, offering tools for navigation, page retrieval, and markdown conversion.

• adamdude828/mcp-browser-use: Facilitates note storage and summarization with a custom URI scheme and prompt-based summarization tool.

• skrapeai/skrape-mcp: Transforms any webpage into clean, LLM-ready Markdown using skrape.ai, ideal for seamless integration with LLMs like Claude and GPT.

• Cam10001110101/mcp-server-browser-use-ollama: Enables AI agents to automate and control web browsers using locally-hosted Ollama models through the Model Context Protocol.

• JovaniPink/mcp-browser-use: Facilitates browser automation through natural language interactions using the browser-use library, built on the MCP protocol.

• cmann50/mcp-chrome-google-search: Facilitates Google search and webpage content extraction via Chrome, integrating with Claude for seamless content fetching.

• PhialsBasement/MCP-Puppeteer-Linux: A Model Context Protocol server enabling browser automation on Linux using Puppeteer, with support for X11 and Wayland display servers.

• tesla0225/chromeextension: Facilitates interaction between Claude and Chrome browser extensions through a dedicated MCP server.

• ryuhei-linkai/mcp-browser: Facilitates natural language-driven browser automation using the Model Context Protocol.

• spences10/mcp-jinaai-reader: Integrates Jina.ai's Reader API with LLMs for efficient web content extraction and analysis.

• Saik0s/mcp-browser-use: AI-driven server for natural language browser automation and web research, leveraging the Model Context Protocol.

• random-robbie/mcp-web-browser: Facilitates secure, headless web browsing and interaction through a flexible API using Playwright.

• tanevanwifferen/usescraper-mcp-server: A TypeScript-based server leveraging the UseScraper API to extract web content into text or markdown formats.

• jitsmaster/WebScrapeMCPServer: Facilitates web crawling tasks with customizable depth and concurrency settings, accessible via MCP.

• bsmi021/mcp-server-webscan: A server for web content scanning and analysis, offering tools for fetching, analyzing, and extracting information from web pages.

• jwaldor/mcp-scrape-copilot: Enables LLMs to automate browser tasks using Puppeteer for web interaction and JavaScript execution.

• adamdude828/playwright_mcp: Facilitates browser automation through Playwright using the Model Context Protocol, enabling seamless integration with MCP-compatible clients.

• blackwhite084/playwright-plus-python-mcp: Enhance browser automation with Playwright tools, offering note storage and summarization capabilities.

• apify/actors-mcp-server: Facilitates seamless interaction with Apify Actors for data extraction and web scraping tasks via the Model Context Protocol.

• hrmeetsingh/mcp-browser-automation: Automate browser interactions using Playwright through a Model Context Protocol server integrated with Claude Desktop.

• tokenizin-agency/mcp-npx-fetch: Fetch and transform web content into multiple formats with high performance and seamless MCP integration.

• pskill9/website-downloader: Facilitates downloading entire websites while preserving structure and converting links for local use.

• tinyfish-io/agentql-mcp: Integrates AgentQL's data extraction capabilities to extract structured data from web pages using the Model Context Protocol.

• JoseRoberts87/spidy: A web crawler MCP server for Claude that stores notes and generates summaries, featuring a web-crawling tool to fetch and save website content.

• apify/mcp-server-rag-web-browser: Facilitates AI agents and LLMs to perform web searches and extract content using the RAG Web Browser Actor.

• mendableai/firecrawl-mcp-server: Enhance web scraping capabilities for LLM clients like Cursor and Claude with Firecrawl's robust MCP server, offering features like JS rendering, batch processing, and smart content filtering.

• rdvo/mcp-server: Enables LLMs to automate browser tasks using Puppeteer and Steel for enhanced web interaction and content extraction.

• steel-dev/steel-mcp-server: Facilitates web navigation and automation for LLMs using Puppeteer and Steel, enabling tasks like searching, form filling, and price comparison.

• executeautomation/mcp-playwright: Automates browser interactions and API testing using Playwright, enabling LLMs to interact with web pages and execute JavaScript in a real browser environment.

• twolven/mcp-server-puppeteer-py: A Python-based MCP server leveraging Playwright for comprehensive browser automation, enabling LLMs to interact with web pages, capture screenshots, and execute JavaScript.

• Automata-Labs-team/MCP-Server-Playwright: Empower LLMs with browser automation capabilities using Playwright for seamless web interaction and JavaScript execution.

• jakedahn/deno2-playwright-mcp-server: Facilitates browser automation through Playwright and Deno 2, enabling LLMs to interact with web pages in a real browser environment.

• Sunwood-ai-labs/duckduckgo-web-search: A TypeScript-based MCP server for Claude Desktop that implements a simple notes system using DuckDuckGo API.

• blazickjp/web-browser-mcp-server: Empower AI models to navigate and extract data from websites using a streamlined MCP interface with advanced content targeting and error management.

• bzsasson/screaming-frog-mcp: Screaming Frog SEO Spider integration for crawling websites, exporting SEO data, and managing crawl storage through AI assistants.

• ExceptionRegret/Kryfto: Open-source web-browsing backend for AI agents with a 42-tool MCP server, REST API for n8n/Zapier/Make, federated multi-engine search, anti-bot stealth, and enterprise infrastructure (Postgres, Redis, BullMQ, MinIO).

• Skyvern-AI/skyvern: Automates browser-based workflows using LLMs and computer vision, with a Playwright-compatible SDK, no-code workflow builder, and MCP server for AI agents to control any website.

• Cap-of-tea/GDD: AI-controlled browser farm — run N isolated Chromium instances with 36 MCP tools for device emulation, screenshots, network/geolocation simulation, and console/network diagnostics. Cross-platform: Windows GUI with video wall + headed/headless on Linux/macOS.

• hshintelligence/agent-scrape: Pay-per-call web scraping for AI agents via the x402 protocol on Base USDC. Six tools (scrape, extract via Groq+Llama 4, screenshot, metadata, session, workflow). No signup, no API keys. Remote MCP at https://agent-scrape.healingsunhaven.workers.dev/mcp. MIT licensed.

• Custodia-Admin/pagebolt-mcp: Provides hosted web capture and browser automation via REST API and MCP. Supports screenshot, PDF, OG image, multi-step sequences, AI-narrated video recording, and page inspection (CSS selectors + element maps for AI agents). Works natively in Claude Desktop, Cursor, and Windsurf.

• opentabs-dev/opentabs: Plugin-based MCP server that gives AI agents access to web applications through the user's authenticated browser session. Chrome extension with 100+ plugins for Slack, Discord, GitHub, Reddit, and more.

• ai-dashboad/flutter-skill: AI-powered E2E testing bridge for any app. Supports Flutter, iOS, Android, Web, Electron, Tauri, KMP, React Native, .NET MAUI.

• kengerlwl/phoneMcp: Enables AI assistants to control Android phones via ADB. Supports zero-config Skill mode (CatPaw/Claude Code) and standard MCP server mode (Claude Desktop/Cursor). Features screen capture, tap/swipe gestures, text input, app management, file transfer, and shell commands.

• achiya-automation/safari-mcp: Native Safari browser automation for AI agents via AppleScript — 80 tools, zero overhead, keeps logins, runs silently in background. macOS only.

• yifanyifan897645/webcheck-mcp: Website health checker MCP server: SEO audit, accessibility scan, broken link detection, and performance analysis.

• yifanyifan897645/deadlink-checker-mcp: Dead link checker MCP server for websites: detects broken links, redirects, and timeouts with detailed reporting.

• searchapi/searchapi-mcp - Real-time SERP and web data for AI assistants: 100+ tools across Google, Bing, Amazon, YouTube, Maps, travel, and social from one remote MCP server.

• skuzadev/wpfpilot-mcp: Automates Windows WPF desktop apps for AI agents using semantic UI Automation selectors, query/act/wait/assert tools, screenshots, workflow recording, and xUnit + FlaUI test generation. Works with Codex, Claude, Cursor, VS Code, and other MCP clients via npx -y @skuzadev/wpfpilot-mcp.

• babelwrap/babelwrap-mcp: HTTP API and MCP server that lets AI agents interact with websites through natural language instead of CSS selectors. Navigate, click, fill forms, and extract data via plain English descriptions.

• massanaRoger/extracto-mcp: Turns any URL plus a JSON schema into validated, typed JSON via the Extracto API. No CSS selectors; missing fields return null instead of guesses. Install: npx -y extracto-mcp.

• ShotAPI: Web screenshot and HTML render MCP server for AI agents. Hosted Streamable HTTP endpoint at https://aiphotoshop.mynatapp.cc/mcp with three tools for quick screenshots, full-control screenshots, and rendering HTML/CSS to images. Free hosted tier requires no signup; local stdio mode supports optional paid-tier keys. MIT.

• PramodDutta/browserbash: Plain-English browser testing for AI agents. Run objectives, Markdown test files, or parallel suites in a real Chrome and get deterministic verdicts back (assertions, exit codes, cost). Tools: run_objective, run_test_file, run_suite. Install: npx browserbash-cli mcp. Free local models via Ollama, no API keys. Apache-2.0.

• raintree-technology/docpull: Python CLI/SDK/MCP server for fetching, caching, searching, validating, and exporting public static and server-rendered web sources as cited Markdown, NDJSON, SQLite, and local context packs. Install: pip install 'docpull[mcp]'.

• protostatis/unbrowser: Lightweight local MCP browser for agents: executes JavaScript, clicks and follows links, fills forms, manages cookies, and returns compact BlockMaps without Chrome.

• Latchshot: Hosted guarded screenshots and PDFs of public pages, returned inline with render and quota diagnostics. Remote: https://latchshot.fly.dev/mcp. Transport: streamable-http. Auth: Bearer API key. Tools: capture_page, get_usage.

• Ghost in the Droid: Control Android devices through ADB with MCP tools for taps, swipes, screenshots, screen trees, app launch, and on-device inference. Install: pip install ghost-in-the-droid.

• hongnoul/hwatu: Visual verification browser for AI coding agents — daemon-based WebKitGTK with real rendering and ~13ms window spawn. Headless or headed per window, DOM eval, screenshots, pixel-diff verification, and live hand-off of a session to a human. Linux/Wayland, single static binary, no Chromium. Install: yay -S hwatu (Arch) or the install script; run hwatu mcp. Stdio transport. MIT.

• bitterdev/scraper-api-mcp: Thin adapter that exposes every scraper-api.com endpoint as its own tool, one per endpoint, so the agent asks for the data instead of driving a browser. Covers public profiles, posts, reels, transcripts, comments and ads across the large social platforms, plus German business directories and real estate portals, returned as consistent JSON. No scraping logic runs locally, calls are forwarded to the HTTPS API. Transport: stdio, or a hosted OAuth endpoint at https://api.scraper-api.com/mcp. Auth: API key. License: MIT. Install: npx -y scraper-api-mcp.

• Crawlora-org/crawlora-mcp: Hosted MCP server for structured public web data. Agents call 790+ tools over 820+ documented REST endpoints across 71 platforms (search/SERP, maps and local, e-commerce, app stores, social, media, finance) and get normalized JSON back instead of raw HTML to parse.

• TestMu AI MCP Server: Hosted MCP server for the TestMu AI (formerly LambdaTest) cloud testing platform. Agents orchestrate HyperExecute runs and generate the YAML for them, triage automation failures from command, network and console logs, explain SmartUI visual regressions, and run WCAG accessibility audits against hosted URLs. Remote: https://mcp.lambdatest.com/mcp. Transport: streamable-http. Auth: OAuth 2.1 with dynamic client registration. Docs: https://www.testmuai.com/support/docs/testmu-mcp-server/.

• VPN.FAIL MCP: A remote MCP service for fetching one public HTTP(S) page as readable Markdown, plain text, or HTML through a rotating proxy pool. Public tools: scrape, usage, and service_status.

• CleanExtract: Hosted Streamable HTTP MCP server that converts public HTML or a public URL into token-dense Markdown. Endpoint: https://extract.getstringer.app/mcp. Price: $0.05 per call via x402 v2 USDC on Base.

• Plainport: Hosted, no-auth Streamable HTTP MCP server for deterministic public-web intake, page fetching, structured extraction, and clean Markdown retrieval for AI agents. Connect: https://plainport.exportitnow.workers.dev/mcp. Docs: https://plainport.exportitnow.workers.dev/docs. Official Registry id dev.workers.exportitnow.plainport/plainport; MIT.

• Pangolin-spg/pangolinfo-mcp: Hosted Streamable HTTP MCP server for live Amazon marketplace intelligence across 20+ marketplaces: product details, reviews, keyword search, category and niche analysis, Best Sellers, New Releases, seller storefronts, SERP features, keyword trends, Maps POI, WIPO trademark search, and US PACER patent litigation lookups. 19 read-only tools. Endpoint: https://mcp.pangolinfo.com/mcp with Authorization: Bearer <PANGOLINFO_API_KEY>. Local bridge: PANGOLINFO_API_KEY=your-key npx -y pangolinfo-amazon-data-mcp; Docker image pangolinfo/amazon-data-mcp:latest. MIT.

• jrpool/kilotest: Ensemble testing of web pages for front-end quality (accessibility, usability, and standards conformity) with 12 rule engines, consolidating ~1300 rules into ~380 issues. Streamable HTTP at https://kilotest.com/mcp; no auth; MIT; docs: https://kilotest.com/llms-full.txt.

• KitchenSink4AI/KitchenSink4Web: Browser automation with budgeted page reads, an honest account of what went unread, structured extraction with named evidence, and a consent ladder for acting. Read-only by default. Install: pip install kitchensink4web. AGPL-3.0.

• AIHawk: AI browser agent that navigates, clicks, types, and reads the real web from plain-English instructions. Ships as an MCP server for Claude Code, Codex, and Gemini CLI, plus a standalone web UI. Install: uvx aihawk (MCP server) or uvx aihawk ui --openrouter-key <key> (web UI). MIT.

• Base HTTP 402 Web Extractor: Autonomous web-to-markdown extraction for LLM context windows, featuring noise removal, token estimation, and on-chain micropayment settlement ($0.05 USDC) on Base L2. Stdio MCP server with get_payment_info and extract_clean_markdown; run from GitHub with npx -y github:psicossz29-netizen/base-http402-extractor-api. Package metadata declares MIT.

• ZennoLab/zennoposter-mcp: Local Windows MCP servers for editing ProjectMaker automation projects, running ZennoPoster tasks, controlling browser instances and ZennoDroid devices. Download the server binaries and create a product API key; the key is passed to the local server, while MCP clients connect to loopback HTTP (ProjectMaker defaults to http://localhost:6207/mcp). Requires a running ZennoPoster 7.9.2+ or ZennoDroid 2.6.1+. Install guide.