Browser Automation MCP Servers
Browser control, page inspection, screenshots, web tests, and automation.

Noetive
io.noetive
Connects an AI editor to Noetive Semantik so agents can publish what they learn into a namespace and find peers' notes by meaning.
oassis — web scraping, crawling and browser control for AI agents
dev.oassisOfficial publisher
Remote MCP server that lets an AI assistant scrape pages, crawl sites, search the web and drive a browser session, billed per credit.

Nova Cast
Published by GitHub account r4v3n-art
Nova Cast lets an AI assistant build, preview and publish interactive components on a paired home TV canvas, and drive them with typed clicks, fills and scripts
Menso
io.mensoOfficial publisher
Menso runs AI users through real tasks on your live website and returns a TRACES score, friction points and a step-by-step replay.

MateMCP
Published by GitHub account vrassouli
Connects a compatible AI chat to your own Windows or macOS computer so it can work on local projects, shells, browsers, and files.

Connections
Published by GitHub account Lunarwerx
A local MCP server that gives an assistant access to AWS, GitHub and connected services without pasted keys, plus local shell and browser control.

Buoy Gg
Published by GitHub account Buoy-gg
Buoy's MCP server lets an AI editor read and change a running React Native, Expo or Flutter app through Buoy's in-app devtools.
oassis — web for agents
dev.oassis
A remote MCP endpoint that lets an assistant read, map and crawl the web and drive a real browser, paid per call without a key or signup.

NativePort
ai.nativeportOfficial publisher
Remote MCP endpoint that puts 30+ web-access and AI APIs — search, scraping, browsers, voice, OCR and LLMs — behind one NativePort key.
Ironfang
com.ironfangOfficial publisher
Ironfang is a remote MCP server that lets an assistant render pages to images and PDFs, audit websites, explain Peppol rules, and run integration test rigs.

Unbrowse
Published by GitHub account unbrowse-ai
Unbrowse lets an assistant read web pages and run site tasks through websites' own first-party APIs, using a hosted cloud browser with saved logins.

Clawdcursor
Published by GitHub account AmrDab
Local MCP server that lets an AI agent read and control the real desktop on Windows, macOS and Linux by acting on UI elements rather than pixels.

Cdpilot
Published by GitHub account mehmetnadir
Lets an assistant drive a local Chrome, Brave, or Chromium browser over raw CDP for navigation, clicking, screenshots, and page inspection.

Glamour
Published by GitHub account gratise
Glamour gives coding agents browser-measured visual debugging and verification tools for checking rendered output.

agentpc
Published by GitHub account pawanpaudel93
Runs resettable Windows, Ubuntu and Arch Linux ARM desktop VMs on an Apple Silicon Mac, letting an AI agent create, drive, screenshot and reset them.
AgentLaunch
Published by GitHub account cracked-collective
AgentLaunch lets an assistant plan and pace a startup or SaaS directory launch, writing per-directory copy and queuing submissions that a Chrome extension fills

Oh My Android
Published by GitHub account ateymoori
Lets an AI agent see and drive an Android emulator or phone via adb: screenshots, UI tree in dp, taps, typing, device settings, logcat and app data.
search-rotation
Published by GitHub account localfoundry
Local MCP server that gives an assistant web search and Markdown page extraction, rotating across several search providers with automatic failover.

lintlab Screenshot Diff
Published by GitHub account lintlab
Remote Apify-hosted MCP server that lets an assistant capture full-page website screenshots and compare two captures as a visual diff.

Usable Browser Agent
Published by GitHub account Savvy-Tech-Sphere
A local MCP server plus Firefox extension that lets an AI agent drive your real, logged-in browser through 53 browser tools.

Screen Browser
com.screenbrowserOfficial publisher
Screen Browser's hosted MCP server, published by screenbrowser.com; credential required; Streamable HTTP.
Scalebrowser
net.scalebrowser
Connects an AI agent to Scalebrowser, a Windows desktop app that gives each agent its own isolated, persistent browser profile for web tasks.

Capturewren: Website Screenshot API for AI Agents (Image + Text)
Published by GitHub account CyberMax-tools
A remote MCP server that screenshots any URL and returns the image plus page status, title and clean text for AI agents.
Browsentic
Published by GitHub account imshaikot
Browsentic Bridge lets an MCP client drive your real, already signed-in browser through an agent CLI running on your own machine.