Google Surf MCP Server

@HarimxChoi

代码仓库

基本信息

分类

搜索

传输方式

stdio

作者

HarimxChoi

提交者

Harim Choi

关于 Google Surf MCP Server

Google search MCP. No API key. One MCP replaces search + fetch + academic-paper extractor.

#google-search #web-search #playwright#pdf-extraction#academic

Google Surf MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

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

Google Surf MCP Server 工具

`content`, `title`, `excerpt`, `length`, `is_pdf`, `page_count`, `extraction_quality`. Failures return `{ error }`, never throw.

概览

What is Google Surf Mcp?

Google Surf Mcp provides Google search as a tool for any MCP client, combining search, URL fetching, and academic PDF extraction into a single server. It requires no API key, no proxies, and no CAPTCHA solver, and is designed for local use with a persistent Chrome profile.

How to use Google Surf Mcp?

Install via npx google-surf-mcp or clone the repository. Register the server in your MCP client’s configuration (e.g., ~/.claude.json for Claude Code). The first tool call auto-bootstraps a warm Chrome profile; you may see Chrome open briefly.

Key features of Google Surf Mcp

No API key, no proxies, no CAPTCHA solver required.
Five tools: search, search_parallel, extract, search_extract, health.
Academic PDF extraction from arxiv, Nature, PubMed, OpenReview, and more.
CAPTCHA recovery with four configurable modes (OS notification, headed, remote debug, fail‑fast).
Sponsored ads and knowledge panels automatically dropped via geometric verification.
Parallel search and extract for efficiency (up to 10 queries, 4‑worker pool).

Use cases of Google Surf Mcp

Search the web and get clean results without API keys.
Fetch and extract article content from URLs (HTML or PDF).
Retrieve academic paper summaries or full texts from supported repositories.
Run multiple search queries in parallel to compare results.
Monitor server health and configuration via the health tool.

常见问题

What happens when a CAPTCHA appears?

A Chrome window opens for a human to solve the CAPTCHA; each solve preserves the Chrome profile's reputation. Four modes are available: OS notification (default), headed mode, remote debug, or fail‑fast in cloud mode.

Does this server require an API key?

No. It uses a persistent Chrome profile with Playwright and stealth techniques, requiring no API key, proxies, or CAPTCHA solver services.

What are the runtime requirements?

Node 18+ and Google Chrome (or Chromium) installed on the system. The first call auto‑bootstraps a warm profile.

How does the server handle SSRF risks?

By default, `extract` blocks private and loopback addresses (`localhost`, `127.0.0.1`, etc.). Set `SURF_ALLOW_PRIVATE=true` to allow them.

How are search results cached?

Search results are cached for 24 hours by default (`SURF_CACHE_TTL_SEARCH_MS`). Set this to `0` to disable caching, or adjust `SURF_CACHE_MAX_ENTRIES` for LRU limits.

搜索分类下的更多 MCP 服务器

Ransack

chancewalker165-dot

Every SERP, every extractor. One MCP: search, page fetch, cited research, transcripts, product prices and permit records for AI agents, with the source URL on every result.

patent-literature-search-mcp

patsnap

A professional tool for searching Patsnap’s global patent and literature databases using natural language, semantic, or keyword queries. It supports precise filtering and retrieves bibliographic data in Markdown format.

3GPP Scout

3gppscout

Hosted 3GPP MCP server: cited search across Rel-15 through Rel-20 TS/TR specs, recent TDocs, company proposals and meeting decisions. OAuth 2.1; no local index required.

VerifyAPI

Circuit Savers LLC

Fact-check any claim with live web sources. Ed25519-signed receipts. Catches hallucinations.

Desearch

Desearch-ai

AI search, X search and web search for AI agents, plus page extraction and X data tools. Bring your own Desearch API key.