Webpawm

Webpawm is an MCP (Model Context Protocol) server that provides web search capabilities with multiple search engine support.

Features

• Multiple Search Engines: Supports SearXNG, Google, Bing, Brave, Bing CN (China), and Arxiv
• Multiple Transport Modes:
  • HTTP/SSE mode (via webpawm web command)
  • Stdio mode (via webpawm or webpawm std command)
• Smart Search: Intelligent query optimization with result aggregation
• Multi-Engine Search: Search across multiple engines simultaneously
• Flexible Configuration: Support for config file (~/.webpawm/config.json) and environment variables
• Access Logging: Built-in slog-based HTTP access logging

Installation

go install github.com/liut/webpawm@latest

Or build from source:

git clone https://github.com/liut/webpawm.git
cd webpawm
go build -o webpawm .

Usage

Stdio Mode (Default)

Run as a local MCP tool:

webpawm

HTTP/SSE Mode

Start the web server:

webpawm web --listen localhost:8087

The server provides two endpoints:

• HTTP: http://localhost:8087/mcp
• SSE: http://localhost:8087/mcp/sse

With URI prefix:

webpawm web --listen localhost:8087 --uri-prefix /api

Endpoints become:

• HTTP: http://localhost:8087/api/mcp
• SSE: http://localhost:8087/api/mcp/sse

MCP Tools

Webpawm provides three MCP tools:

web_search

Search the web using a single search engine.

Parameter	Type	Description
query	string	The search query (required)
engine	string	Search engine to use (optional)
max_results	integer	Maximum results (1-50, default: 10)
language	string	Language code (e.g., 'en', 'zh')
arxiv_category	string	Arxiv category for academic papers

multi_search

Search across multiple search engines simultaneously.

Parameter	Type	Description
query	string	The search query (required)
engines	array	List of search engines (optional)
max_results_per_engine	integer	Max results per engine (1-20, default: 5)

smart_search

Intelligently search with query optimization and result aggregation.

Parameter	Type	Description
question	string	User's question or search intent (required)
search_depth	string	'quick', 'normal', or 'deep'
include_academic	boolean	Include academic papers from Arxiv

Configuration

Config File

Create ~/.webpawm/config.json:

{
  "searchxng_url": "https://searchx.ng",
  "google_api_key": "your-api-key",
  "google_cx": "your-search-engine-id",
  "bing_api_key": "your-bing-api-key",
  "brave_api_key": "your-brave-api-key",
  "max_results": 10,
  "default_engine": "searchxng",
  "listen_addr": "localhost:8087",
  "uri_prefix": "",
  "log_level": "info"
}

Environment Variables

Variable	Description
WEBPAWM_SEARCHXNG_URL	SearXNG base URL
WEBPAWM_GOOGLE_API_KEY	Google Custom Search API key
WEBPAWM_GOOGLE_CX	Google Search Engine ID
WEBPAWM_BING_API_KEY	Bing Search API key
WEBPAWM_BRAVE_API_KEY	Brave Search API key
WEBPAWM_MAX_RESULTS	Default max results
WEBPAWM_DEFAULT_ENGINE	Default search engine
WEBPAWM_LISTEN_ADDR	HTTP listen address
WEBPAWM_URI_PREFIX	URI prefix for endpoints
WEBPAWM_LOG_LEVEL	Log level: debug, info, warn, error

Priority

Configuration priority (highest to lowest):

1. CLI flags
2. Environment variables
3. Config file
4. Default values

License

MIT