Website Scraper

A command-line tool and MCP server for scraping websites and converting HTML to Markdown.

Features

• Extracts meaningful content from web pages using Mozilla's Readability library (the same engine used in Firefox's Reader View)
• Converts clean HTML to high-quality Markdown with TurndownService
• Securely handles HTML by removing potentially harmful script tags
• Works as both a command-line tool and an MCP server
• Supports direct conversion of local HTML files to Markdown

Installation

# Install dependencies
npm install

# Build the project
npm run build

# Optionally, install globally
npm install -g .

Usage

CLI Mode

# Print output to console
scrape https://example.com

# Save output to a file
scrape https://example.com output.md

# Convert a local HTML file to Markdown
scrape --html-file input.html

# Convert a local HTML file and save output to a file
scrape --html-file input.html output.md

# Show help
scrape --help

# Or run via npm script
npm run start:cli -- https://example.com

MCP Server Mode

This tool can be used as a Model Context Protocol (MCP) server:

# Start in MCP server mode
npm start

Code Structure

• src/index.ts - Core functionality and MCP server implementation
• src/cli.ts - Command-line interface implementation
• src/data_processing.ts - HTML to Markdown conversion functionality

API

The tool exports the following functions:

// Scrape a website and convert to Markdown
import { scrapeToMarkdown } from './build/index.js';

// Convert HTML string to Markdown directly
import { htmlToMarkdown } from './build/data_processing.js';

async function example() {
  // Web scraping
  const markdown = await scrapeToMarkdown('https://example.com');
  console.log(markdown);
  
  // Direct HTML conversion
  const html = 'Hello WorldThis is bold text.';
  const md = htmlToMarkdown(html);
  console.log(md);
}

License

ISC