MCP 服务器目录

Self-hosted stealth browser for AI agents: Firefox engine + Rust MCP runtime.

Speech, transcription, voice agents, Trace, Recap, dubbing and narration with browser OAuth.

Debug PWAs in your real browser via MCP: service-worker, cache, installability & framework state.

Drive Chrome with filesystem commands (ls, cd, grep, click, type). One MCP tool, multi-agent.

Free music for your coding agent. Play songs, mood stations and Spotify links in your browser.

Multi-session browser MCP — 15 parallel isolated browsers, 37 tools, stealth, self-improvement.

Undetected anti-bot Chrome automation. 70 tools, parallel isolated browsers.

Cast YouTube videos and media to a Chromecast or DLNA device with catt, and control playback.

WCAG 2.1 AA accessibility scanner (Playwright + axe-core) with grounded, browser-verified fixes.

AI-native web scraper: scrape, crawl and map any site to clean markdown over stdio. MIT-licensed.

Browser-based QA for AI-built software. Test pages with real browsers via agents.

Local Playwright browser through residential proxies — agent picks country + sticky identity

MCP server for Mint — AI-powered QA that runs your app in a real browser on every PR.

Native Safari browser automation for AI agents — 98 tools, zero Chrome overhead.

Zenrows MCP server — Fetch, Extract, Batch, and Browser Sessions for AI coding assistants

Pay-per-call browser rendering and x402 marketplace ranking for AI agents, billed via USDC.

MCP server for JavaScript reverse engineering and browser debugging.

MCP server for JavaScript reverse engineering and browser debugging.

Give any LLM its own computer — Docker sandboxes with bash, browser, docs, and sub-agents

Use Reddit, Zhihu, X and WhatsApp through your signed-in browser with compact, on-demand MCP tools.

Revision-bound quality planning, execution, evidence, proof, and Playwright authoring.

MCP server for Island Browser's Admin Actions, Audit, Compromised Credential, and Device APIs.

Drive your real, logged-in Chrome from any MCP client — even from a remote machine, no inbound port.

User story map + kanban with a live browser UI; AI agents and humans plan the same backlog via MCP.