Web Tools Mcp

io.github.mlawsonkingPublished by GitHub account mlawsonkingv1.3.0Updated Oct 24, 2026

10 no-LLM web tools: URL to Markdown, metadata, JSON-LD, MX, scrape, RSS, DNS, RDAP, SSL, HTTP.

VerifiedSTDIODesktop onlyWeb Search & ScrapingDeveloper Tools

Overview

AI-generated overview

A local MCP server of ten deterministic web utilities: page-to-Markdown, metadata, scraping, feeds, DNS, RDAP, SSL and HTTP checks.

What you can do
  • Read pages as clean Markdown and pull titles, descriptions, preview images and favicons
  • Extract JSON-LD, OpenGraph and Twitter card data, or scrape fields with CSS selectors
  • Fetch RSS/Atom feeds as JSON and validate email addresses with a live MX lookup
  • Run DNS lookups with SPF/DMARC detection, RDAP domain registration details, TLS certificate checks and HTTP redirect/header inspection
When to use it
Useful when an assistant needs to read or extract web content for research or RAG, or to run quick domain, DNS, certificate and HTTP diagnostics without an LLM or API key.
Requirements
Runs locally over stdio, typically launched with npx from the npm package web-tools-mcp, so Node.js and network access are needed. No authentication or environment variables are declared; the README mentions an optional TOOLS_API_URL for self-hosting.
Before you install
The README states the server contacts a remote API and checks a rules feed about once a day, applying signed rule updates automatically; it says the request carries only the surface name and installed rules version. Updates can be disabled with --offline, AGENT_GUARDS_NO_FEED=1 or a config setting. Tools fetch arbitrary URLs and send them to that API.

Installation

In SourceWeft

  1. Open Web Tools Mcp 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

web-tools-mcp

An MCP server with a set of web utilities for AI agents: read a page as Markdown, pull link metadata, scrape with CSS selectors, parse feeds, and run DNS, RDAP, SSL, and HTTP checks. Deterministic, no LLM, and no API keys needed for the free tier.

Install

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

Tools

  • read_url: fetch a page and return the main content as clean Markdown, with the nav and ads stripped out. Good for RAG.
  • unfurl_url: get a URL's title, description, preview image, site name, and favicon.
  • validate_email: syntax check plus a live MX lookup, with disposable, role, and free-provider detection.
  • extract_web: scrape a page with CSS selectors and get back the fields you asked for.
  • get_feed: fetch an RSS or Atom feed and return the items as JSON.
  • dns_lookup: DNS records (A, AAAA, MX, NS, TXT, and the rest) plus SPF and DMARC detection.
  • domain_info: registration details via RDAP, including domain age, registrar, and expiry.
  • ssl_check: inspect a host's TLS certificate: issuer, validity window, days until expiry, and SANs.
  • http_inspect: final status, the full redirect chain, response headers, and a security-header report.
  • structured_data: extract JSON-LD, OpenGraph, and Twitter card data from a page.

Rule updates

About once a day this server asks the rules feed whether there is a newer ruleset, and applies it if there is. The request carries two things: which surface asked, which here is facade, and the rules version already installed. No machine id, no user id, no file names, nothing you scanned.

Bundles are signed with Ed25519 and verified against a public key compiled into this package, so it does not matter which mirror served one. A bundle that fails its signature, its schema, or its ReDoS check is discarded and the rules you already had stay in place. --offline turns updates off, as does AGENT_GUARDS_NO_FEED=1 or {"feed": false} in ~/.agent-guards/config.json. The bundle format and how to verify one yourself: https://github.com/mlawsonking/MCP/blob/main/rules/README.md

It calls the API at https://agent-tools-api.vercel.app (set TOOLS_API_URL if you self-host). One of six agent tools and guards in this repo: package-guard, agent-firewall, payment-guard, email-guard, code-guard, and web-tools. MIT.

Source: agent-tools-mcp/README.md at commit 91f3fa7

Tools

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

Version history

1
  1. v1.3.0LatestSep 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