Wick MCP Server

@wickproject

代码仓库

基本信息

分类

浏览器自动化

传输方式

stdio

作者

wickproject

提交者

Myles Horton

关于 Wick MCP Server

Browser-grade web access for AI agents. Fetches any webpage using Chrome's actual network stack — identical TLS fingerprint to a real browser. Bypasses Cloudflare, Akamai, and other anti-bot systems that block standard HTTP clients. Returns clean markdown. Built by the creator of

#web #fetch #browser #scraping #anti-bot #markdown #search #cloudflare #mcp #ai-agent

Wick 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": {
    "wick": {
      "command": "wick",
      "args": [
        "serve",
        "--mcp"
      ]
    }
  }
}

Wick MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表，不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Wick?

Wick is an MCP server that uses Chrome’s actual network stack to give AI agents browser-grade web access. It runs locally from your own IP, returns clean markdown, and bypasses anti-bot protections (Cloudflare, Akamai, PerimeterX, AWS WAF) with a 100% success rate on tested sites.

How to use Wick?

Install via Homebrew (macOS), apt (Linux), or npm (any platform), then run wick setup to auto-configure MCP clients like Claude Code, Cursor, etc. Use tools such as wick_fetch, wick_crawl, wick_map, wick_search, and wick_download directly from your agent. Wick also provides a local HTTP API server accessible via wick serve --api.

Key features of Wick

wick_fetch – Fetch any URL and get clean, LLM-friendly markdown.
wick_crawl – Crawl a website, following same-domain links up to a configurable depth.
wick_map – Discover all URLs on a site via sitemap.xml or link traversal.
wick_search – Perform web searches from the command line or MCP.
wick_download – Download video/audio from Reddit, YouTube, Twitter, and 1000+ sites.
wick_session – Clear cookies and session data to start fresh.

Use cases of Wick

Fetching a paywalled or bot-protected news article for summarization.
Crawling documentation sites to feed an agent multi-page context.
Mapping all URLs of a website for sitemap generation or auditing.
Downloading media from social platforms for downstream processing.
Searching the web and then reading full result pages in markdown.

常见问题

How does Wick bypass anti-bot protections?

Wick uses Chrome’s actual network stack (Cronet – the same BoringSSL and HTTP/2 implementation in real Chrome), so the TLS fingerprint is indistinguishable from a real browser. No simulation or header spoofing is needed.

What platforms are supported?

Wick supports macOS (Homebrew), Linux (apt), and any platform via npm. The binary can also be built from source using `cargo build --release`.

Does Wick run locally?

Yes, Wick runs entirely locally from your own IP. Your data stays on your machine, and there are no cloud dependencies.

What is Wick Pro?

Wick Pro adds a real Chromium browser engine with stealth patches for JavaScript rendering and advanced anti-detection. It costs $20/month flat with no per-request billing.

How does Wick compare to Firecrawl?

In testing, Wick achieved a 100% anti‑bot success rate vs 83% for Firecrawl. Wick runs locally (Firecrawl is cloud‑based) and offers unlimited free usage versus Firecrawl’s 500‑page free tier.

浏览器自动化分类下的更多 MCP 服务器

Nullprint

nullprintai

Let Claude Code, Claude Desktop, OpenAI Codex, Kimi and any MCP-capable AI agent create, launch, check and drive browser profiles in Nullprint, the AI antidetect browser. Install with one command.

Fetch

modelcontextprotocol

Model Context Protocol Servers

88.1K

Puppeteer MCP Server

merajmehrabi

This MCP server provides browser automation capabilities through Puppeteer, allowing interaction with both new browser instances and existing Chrome windows.

474