Web

io.github.rog0xPublished by GitHub account rog0xv1.0.1Updated Oct 25, 2026

Web scraping, search, monitoring, and HTML-to-markdown for AI agents

VerifiedSTDIODesktop onlyWeb Search & ScrapingKnowledge & Memory

Overview

AI-generated overview

A local MCP server that gives an assistant web scraping, DuckDuckGo search, site monitoring and HTML-to-markdown tools.

What you can do
Extract structured content from a URL: title, headings, paragraphs, links, images and metadata, or from several URLs in parallel via web_multi_extract (web_extract, web_multi_extract). Search the web through DuckDuckGo without an API key (web_search). Check whether a site is up, measure response time and read HTTP headers (web_monitor). Convert a URL or raw HTML into clean markdown (html_to_markdown).
When to use it
Useful when an assistant needs to read or summarize live web pages, gather search results, batch-fetch several pages at once, or check that an endpoint is responding. A lightweight alternative to headless-browser setups for plain content extraction.
Requirements
Runs locally over stdio via npx or a global npm install; needs Node.js 18 or newer and network access to the target sites and DuckDuckGo. No API keys or environment variables are declared.
Before you install
It fetches arbitrary URLs supplied by the assistant, so requests go to third-party sites and those sites see the traffic. Extracted page content is untrusted input that may carry prompt injection. Search relies on DuckDuckGo's public interface, which can rate-limit or change. No credentials are requested.

Installation

In SourceWeft

  1. Open Web in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

MCP Web Tools

A powerful MCP (Model Context Protocol) server that gives AI agents web superpowers. Extract content, search the web, monitor sites, and convert HTML to markdown — all through a clean tool interface.

Tools

ToolDescription
web_extractExtract structured content from any URL (title, headings, paragraphs, links, images, metadata)
web_searchSearch the web via DuckDuckGo — no API key needed
web_monitorCheck if a site is up, measure response time, get HTTP headers
html_to_markdownConvert any URL or HTML to clean markdown
web_multi_extractExtract content from multiple URLs in parallel

Installation

For Claude Code

Add to your Claude Code MCP settings (~/.claude/settings.json):

json
{  "mcpServers": {    "web-tools": {      "command": "npx",      "args": ["@rog0x/mcp-web-tools"]    }  }}

For Claude Desktop

Add to claude_desktop_config.json:

json
{  "mcpServers": {    "web-tools": {      "command": "npx",      "args": ["@rog0x/mcp-web-tools"]    }  }}

Manual Installation

bash
npm install -g @rog0x/mcp-web-toolsmcp-web-tools

Usage Examples

Extract content from a webpage

Use web_extract to get the main content from https://example.com

Search the web

Use web_search to find "best TypeScript frameworks 2026"

Monitor a website

Use web_monitor to check if https://api.example.com is responding

Convert page to markdown

Use html_to_markdown to convert https://docs.example.com/guide to readable markdown

Batch extraction

Use web_multi_extract to get content from these URLs: [url1, url2, url3]

Features

  • No API keys required — uses DuckDuckGo for search, direct HTTP for everything else
  • Fast — parallel extraction, 15s timeout per request
  • Clean output — strips scripts, styles, ads; extracts meaningful content
  • Lightweight — minimal dependencies, no headless browser needed
  • Production-ready — proper error handling, timeouts, input validation

Requirements

  • Node.js 18+

License

MIT — rog0x

Source: README.md at commit 49a918e

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v1.0.1LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts