wet-mcp 2.15.0b1

Open-source MCP Server for web search, extract, crawl, academic research, and library docs with embedded SearXNG

WET - Web Extended Toolkit MCP Server

mcp-name: io.github.n24q02m/wet-mcp

Open-source MCP Server for web search, content extraction, library docs & multimodal analysis.

Features

• Web Search -- Embedded SearXNG metasearch (Google, Bing, DuckDuckGo, Brave) with filters, semantic reranking, query expansion, and snippet enrichment
• Academic Research -- Search Google Scholar, Semantic Scholar, arXiv, PubMed, CrossRef, BASE
• Library Docs -- Auto-discover and index documentation with FTS5 hybrid search, HyDE-enhanced retrieval, and version-specific docs
• Content Extract -- Clean content extraction (Markdown/Text), structured data extraction (LLM + JSON Schema), batch processing (up to 50 URLs), deep crawling, site mapping
• Local File Conversion -- Convert PDF, DOCX, XLSX, CSV, HTML, EPUB, PPTX to Markdown
• Media -- List, download, and analyze images, videos, audio files
• Anti-bot -- Stealth mode bypasses Cloudflare, Medium, LinkedIn, Twitter
• Zero Config -- Built-in local Qwen3 embedding + reranking, no API keys needed. Optional cloud providers (Jina AI, Gemini, OpenAI, Cohere)
• Sync -- Cross-machine sync of indexed docs via rclone (Google Drive, S3, Dropbox)

Quick Start

Claude Code Plugin (Recommended)

claude plugin add n24q02m/wet-mcp

MCP Server

Python 3.13 required -- Python 3.14+ is not supported due to SearXNG incompatibility. You must specify --python 3.13 when using uvx.

On first run, the server automatically installs SearXNG, Playwright chromium, and starts the embedded search engine.

Option 1: uvx

{
  "mcpServers": {
    "wet": {
      "command": "uvx",
      "args": ["--python", "3.13", "wet-mcp@latest"],
      "env": {
        // -- optional: cloud embedding + reranking (Jina AI recommended)
        "API_KEYS": "JINA_AI_API_KEY:jina_...",
        // -- or: "API_KEYS": "GOOGLE_API_KEY:AIza...,COHERE_API_KEY:co-...",
        // -- without API_KEYS, uses built-in local Qwen3 ONNX models (CPU, ~570MB first download)
        // -- optional: LiteLLM Proxy (production, selfhosted gateway)
        // "LITELLM_PROXY_URL": "http://10.0.0.20:4000",
        // "LITELLM_PROXY_KEY": "sk-your-virtual-key",
        // -- optional: higher rate limits for docs discovery (60 -> 5000 req/hr)
        "GITHUB_TOKEN": "ghp_...",
        // -- optional: restrict local file conversion to specific directories
        // "CONVERT_ALLOWED_DIRS": "/home/user/docs,/tmp/uploads",
        // -- optional: sync indexed docs across machines via rclone
        "SYNC_ENABLED": "true",                    // default: false
        "SYNC_INTERVAL": "300"                     // auto-sync every 5min (0 = manual only)
      }
    }
  }
}

Option 2: Docker

{
  "mcpServers": {
    "wet": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "--name", "mcp-wet",
        "-v", "wet-data:/data",
        "-e", "API_KEYS",
        "-e", "GITHUB_TOKEN",
        "-e", "SYNC_ENABLED",
        "-e", "SYNC_INTERVAL",
        "n24q02m/wet-mcp:latest"
      ],
      "env": {
        "API_KEYS": "JINA_AI_API_KEY:jina_...",
        "GITHUB_TOKEN": "ghp_...",
        "SYNC_ENABLED": "true",
        "SYNC_INTERVAL": "300"
      }
    }
  }
}

Pre-install (optional)

Use the setup MCP tool to warmup models and install dependencies:

# Via MCP tool call (recommended):
setup(action="warmup")

# With cloud embedding configured, warmup validates API keys
# and skips local model download if cloud models are available.

The warmup action pre-downloads SearXNG, Playwright, and embedding/reranker models (~1.1GB total) so the first real connection does not timeout.

Sync setup

Sync is fully automatic. Just set SYNC_ENABLED=true and the server handles everything:

1. First sync: rclone is auto-downloaded, a browser opens for OAuth authentication
2. Token saved: OAuth token is stored locally at ~/.wet-mcp/tokens/ (600 permissions)
3. Subsequent runs: Token is loaded automatically -- no manual steps needed

For non-Google Drive providers, set SYNC_PROVIDER and SYNC_REMOTE:

{
  "SYNC_ENABLED": "true",
  "SYNC_PROVIDER": "dropbox",
  "SYNC_REMOTE": "dropbox"
}

Tools

Tool	Actions	Description
search	search, research, docs, similar	Web search (with filters, reranking, expand/enrich), academic research, library docs (HyDE), find similar
extract	extract, batch, crawl, map, convert, extract_structured	Content extraction, batch processing (up to 50 URLs), deep crawling, site mapping, local file conversion, structured data extraction (JSON Schema)
media	list, download, analyze	Media discovery, download, and analysis
config	status, set, cache_clear, docs_reindex	Server configuration and cache management
setup	warmup, setup_sync	Pre-download models, configure cloud sync
help	--	Full documentation for any tool

Configuration

Variable	Required	Default	Description
API_KEYS	No	--	LLM API keys for SDK mode (format: ENV_VAR:key,...). Enables cloud embedding + reranking
LITELLM_PROXY_URL	No	--	LiteLLM Proxy URL. Enables proxy mode
LITELLM_PROXY_KEY	No	--	LiteLLM Proxy virtual key
GITHUB_TOKEN	No	auto-detect	GitHub token for docs discovery (60 -> 5000 req/hr). Auto-detected from gh auth token
EMBEDDING_BACKEND	No	auto-detect	litellm (cloud) or local (Qwen3). Auto: API_KEYS -> litellm, else local
EMBEDDING_MODEL	No	auto-detect	LiteLLM embedding model name
EMBEDDING_DIMS	No	0 (auto=768)	Embedding dimensions
RERANK_ENABLED	No	true	Enable reranking after search
RERANK_BACKEND	No	auto-detect	litellm or local. Auto: Cohere/Jina key -> litellm, else local
RERANK_MODEL	No	auto-detect	LiteLLM rerank model name
RERANK_TOP_N	No	10	Return top N results after reranking
LLM_MODELS	No	gemini/gemini-3-flash-preview	LiteLLM model for media analysis
WET_AUTO_SEARXNG	No	true	Auto-start embedded SearXNG subprocess
WET_SEARXNG_PORT	No	41592	SearXNG port
SEARXNG_URL	No	http://localhost:41592	External SearXNG URL (when auto disabled)
SEARXNG_TIMEOUT	No	30	SearXNG request timeout in seconds
CONVERT_MAX_FILE_SIZE	No	104857600	Max file size for local conversion in bytes (100MB)
CONVERT_ALLOWED_DIRS	No	--	Comma-separated paths to restrict local file conversion
CACHE_DIR	No	~/.wet-mcp	Data directory for cache, docs, downloads
DOCS_DB_PATH	No	~/.wet-mcp/docs.db	Docs database location
DOWNLOAD_DIR	No	~/.wet-mcp/downloads	Media download directory
TOOL_TIMEOUT	No	120	Tool execution timeout in seconds (0=no timeout)
WET_CACHE	No	true	Enable/disable web cache
SYNC_ENABLED	No	false	Enable rclone sync
SYNC_PROVIDER	No	drive	rclone provider type (drive, dropbox, s3, etc.)
SYNC_REMOTE	No	gdrive	rclone remote name
SYNC_FOLDER	No	wet-mcp	Remote folder name
SYNC_INTERVAL	No	300	Auto-sync interval in seconds (0=manual)
LOG_LEVEL	No	INFO	Logging level

Embedding & Reranking

Both embedding and reranking are always available -- local models are built-in and require no configuration.

• Jina AI (recommended): A single JINA_AI_API_KEY enables both embedding and reranking
• Embedding priority: Jina AI > Gemini > OpenAI > Cohere. Local Qwen3 fallback always available
• Reranking priority: Jina AI > Cohere. Local Qwen3 fallback always available
• GPU auto-detection: CUDA/DirectML auto-detected, uses GGUF models for better performance
• All embeddings stored at 768 dims. Switching providers never breaks the vector table

LLM Configuration (3-Mode Architecture)

Priority	Mode	Config	Use case
1	Proxy	LITELLM_PROXY_URL + LITELLM_PROXY_KEY	Production (selfhosted gateway)
2	SDK	API_KEYS	Dev/local with direct API access
3	Local	Nothing needed	Offline, embedding/rerank only (no LLM)

SearXNG Configuration (2-Mode)

Mode	Config	Description
Embedded (default)	WET_AUTO_SEARXNG=true	Auto-installs and manages SearXNG as subprocess
External	WET_AUTO_SEARXNG=false + SEARXNG_URL=http://host:port	Connects to pre-existing SearXNG instance

Build from Source

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

Compatible With

Also by n24q02m

Server	Description
mnemo-mcp	Persistent AI memory with hybrid search and cross-machine sync
better-notion-mcp	Markdown-first Notion API with 9 composite tools
better-email-mcp	Email (IMAP/SMTP) with multi-account and auto-discovery
better-godot-mcp	Godot Engine 4.x with 18 tools for scenes, scripts, and shaders
better-telegram-mcp	Telegram dual-mode (Bot API + MTProto) with 6 composite tools
better-code-review-graph	Knowledge graph for token-efficient code reviews

Contributing

See CONTRIBUTING.md.

License

MIT -- See LICENSE.