google-surf-mcp

✨Anti-Bot Search MCP: No API Key✨

English | 한국어

Demo only. Actual searches run headless by default (no visible browser). Set SURF_HEADLESS=false to make Chrome visible like in the clip above.

Google search MCP. No API key. Just works.

• ✅ Actually works (tested 6 free Google search MCPs, all failed)
• ✅ Search + URL extract in one MCP (replaces the usual search MCP + fetch MCP combo)
• ✅ 5 tools: search / search_parallel / extract / search_extract / health
• ✅ No API key, no proxies, no solver
• ✅ Auto CAPTCHA recovery (Chrome opens, human solves once, call retries)
• ✅ SSRF guard on extract (blocks localhost, private IPs, AWS metadata by default)

What

Plug it into any MCP client and you get Google search as a tool.

No CAPTCHA solver. When CAPTCHA fires on any tool, a Chrome window opens for a human to solve. Each solve preserves the profile's reputation with Google. Built for sustainable, ethical use.

One-time install needs a ~1s profile warm-up (see Install).

Designed for local use. For headless / serverless environments set SURF_CLOUD_MODE=true (fail-fast on CAPTCHA, worker pool disabled).

Numbers

Measured on a workstation with a 1Gb/s connection.

Stack

• Playwright + persistent Chrome profile
• playwright-extra stealth as a cascade fallback tier
• Resource-blocked images / media / fonts for speed
• One-shot profile bootstrap before first run
• Mozilla Readability + Turndown for article extraction

Install

Requires Node 18+ and Google Chrome (or Chromium) on the system.

npx google-surf-mcp   # actual MCP - register in client config

Or local clone:

git clone https://github.com/HarimxChoi/google-surf-mcp
cd google-surf-mcp
npm install
npm run bootstrap

bootstrap opens a Chrome window. Run one Google search in it. Close. Profile is now warm.

Override paths if needed:

CHROME_PATH=/path/to/chrome SURF_TZ=America/New_York npm run bootstrap

Use with Claude Code

Paste this into your ~/.claude.json:

{
  "mcpServers": {
    "google-surf": {
      "command": "npx",
      "args": ["-y", "google-surf-mcp"]
    }
  }
}

Restart Claude Code. Done. search, search_parallel, extract, search_extract, health are now available.

For other MCP clients, use the same JSON shape in their config file.

Local clone variant:

{
  "mcpServers": {
    "google-surf": {
      "command": "node",
      "args": ["/abs/path/to/google-surf-mcp/build/index.js"]
    }
  }
}

Tools

• search(query, limit?) - single query, ~1.5s. Returns title / url / snippet. Sponsored ads filtered out. Results cached 24h (SURF_CACHE_TTL_SEARCH_MS=0 to bypass).
• search_parallel(queries[], limit?) - pool of 4, max 10 queries per call.
• extract(url, max_chars?) - fetch a URL, return article markdown (Readability with text fallback). Failures return { error }, never throw.
• search_extract(query, limit?, max_chars?) - search + parallel extract in one call. Returns SERP results enriched with full article content. Per-page failures are isolated.
• health() - server status: cascade mode, rate-limiter usage, cache size, config. Call it if searches start failing or retu

... View full README on GitHub