wet-mcp 3.1.0b1

Open-source MCP Server for web search, extract, and crawl with embedded SearXNG

WET - Web ExTract MCP Server

Open-source MCP Server for web scraping & multimodal extraction.

Features

• Web Search - Search via embedded SearXNG (metasearch: Google, Bing, DuckDuckGo, Brave)
• Content Extract - Extract clean content (Markdown/Text)
• Deep Crawl - Crawl multiple pages from a root URL with depth control
• Site Map - Discover website URL structure
• Media - List and download images, videos, audio files
• Anti-bot - Stealth mode bypasses Cloudflare, Medium, LinkedIn, Twitter

---

Quick Start

Prerequisites

• Python 3.13+ (or use uvx)

Add to mcp.json

uvx (Recommended)

{
  "mcpServers": {
    "wet": {
      "command": "uvx",
      "args": ["wet-mcp@latest"],
      "env": {
        "API_KEYS": "GOOGLE_API_KEY:AIza..."
      }
    }
  }
}

That's it! On first run:

1. Automatically installs SearXNG from GitHub
2. Automatically installs Playwright chromium + system dependencies
3. Starts embedded SearXNG subprocess
4. Runs the MCP server

Docker

{
  "mcpServers": {
    "wet": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "-e", "API_KEYS", "n24q02m/wet-mcp:latest"],
      "env": {
        "API_KEYS": "GOOGLE_API_KEY:AIza..."
      }
    }
  }
}

Without uvx

pip install wet-mcp
wet-mcp

---

Tools

Tool	Actions	Description
web	search, extract, crawl, map	Web operations
media	list, download, analyze	Media discovery & download
help	-	Full documentation

Usage Examples

{"action": "search", "query": "python web scraping", "max_results": 10}
{"action": "extract", "urls": ["https://example.com"]}
{"action": "crawl", "urls": ["https://docs.python.org"], "depth": 2}
{"action": "map", "urls": ["https://example.com"]}
{"action": "list", "url": "https://github.com/python/cpython"}
{"action": "download", "media_urls": ["https://example.com/image.png"]}

---

Configuration

Variable	Default	Description
WET_AUTO_SEARXNG	true	Auto-start embedded SearXNG subprocess
WET_SEARXNG_PORT	8080	SearXNG port
SEARXNG_URL	http://localhost:8080	External SearXNG URL (when auto disabled)
API_KEYS	-	LLM API keys for media analysis
LOG_LEVEL	INFO	Logging level

LLM Configuration (Optional)

For media analysis (images, videos, audio), configure API keys:

API_KEYS=GOOGLE_API_KEY:AIza...
LLM_MODELS=gemini/gemini-3-flash-preview

---

Architecture

┌─────────────────────────────────────────────────────────┐
│                    MCP Client                           │
│            (Claude, Cursor, Windsurf)                   │
└─────────────────────┬───────────────────────────────────┘
                      │ MCP Protocol
                      ▼
┌─────────────────────────────────────────────────────────┐
│                   WET MCP Server                        │
│  ┌──────────┐  ┌──────────┐  ┌──────────────────────┐   │
│  │   web    │  │  media   │  │        help          │   │
│  │ (search, │  │ (list,   │  │  (full documentation)│   │
│  │ extract, │  │ download,│  └──────────────────────┘   │
│  │ crawl,   │  │ analyze) │                             │
│  │ map)     │  └────┬─────┘                             │
│  └────┬─────┘       │                                   │
│       │             │                                   │
│       ▼             ▼                                   │
│  ┌──────────┐  ┌──────────┐                             │
│  │ SearXNG  │  │ Crawl4AI │                             │
│  │(embedded)│  │(Playwright)│                           │
│  └──────────┘  └──────────┘                             │
└─────────────────────────────────────────────────────────┘

---

Build from Source

git clone https://github.com/n24q02m/wet-mcp
cd wet-mcp

# Setup (requires mise: https://mise.jdx.dev/)
mise run setup

# Run
uv run wet-mcp

Docker Build

docker build -t n24q02m/wet-mcp:latest .

Requirements: Python 3.13+

---

Contributing

See CONTRIBUTING.md

License

MIT - See LICENSE