Kernel MCP Server
MCP serverHosted MCP server for running cloud browsers and managing Kernel apps from an agent
Not checked yet- Remote
- Open source
- Browser automation
Hosted MCP server for running cloud browsers and managing Kernel apps from an agent
Not checked yetBrowser-based website checks that coding agents run after each change, with evidence
Not checked yet
Screenshots and PDFs of any web page with cookie banners and ads stripped out
Not checked yetRead-only Playwright CI results for agents: new versus known failures and flaky tests
Runs locallyScreenshot, PDF and Open Graph image tools for agents, with or without an API key
Not checked yetDrives your own signed-in Chrome or Edge from an MCP client to finish web tasks
Not checked yetDesign guidance and headless Chromium QA for AI coding agents building front ends
Runs locallyManage SoftGlaze Browser profiles, proxies and macros on your PC. Every change asks you…
Runs locallyPlaywright MCP-compatible browser automation running on CloakBrowser Chromium
Not checked yetFail-closed browser for agents: navigate, extract, fill forms, human handoff, evidence…
Runs locallyHosted browser MCP connector that opens pages, searches the web and reads PDFs for Claude
Not checked yetGive AI agents controlled access to real Android phones, with approvals and an audit log
Runs locallyScreenshot API for AI agents: any public page as an image or PDF from 227 countries
Not checked yetTurns websites into callable tools so agents skip the browser, with a cloud fallback
Not checked yetDrive Veilus browser profiles, pages, scripts and schedules from an AI assistant
Runs locallyLet Claude Code, Codex or Cursor click, type and read the screen on your computer
Runs locallyCheck pages for responsive design problems and screenshot them on real device sizes
Runs locallyApprove agent actions from your phone; sandboxed browser vault with saved logins
Runs locallyLocal QA tools for coding agents: scan a page, inspect the UI, generate a Playwright repro
Runs locallyBudgeted web page reading for AI assistants, read-only until you allow clicking
Runs locallyGive agents a real Chrome that renders JavaScript: Markdown, screenshots, Lighthouse
Not checked yetGive any agent your signed-in Mac browser, with a confirmation step before risky actions
Runs locallyPre-launch website QA in about 60 seconds using real Chromium, payable by agents in USDC
Not checked yetOne MCP gateway for QA work: Playwright, files, GitHub, Jira and databases together
Runs locally
Verify an isolated WPE WebKit runtime with native input, DOM and screenshot evidence
Runs locally
Turn a public URL into a screenshot, PDF or text extract with a render receipt
Not checked yetLet an agent drive a hosted Android phone: tap, type, install APKs, read crash logs
Runs locallyTurns websites and signed-in accounts into callable APIs, run in the cloud or locally
Not checked yetLaunch and check mobile device simulations from an AI agent in your editor
Not checked yetRead-only website smoke tests in a real browser, run by your AI agent over MCP
Not checked yet