Rubber Duck Scraper

Python MCP server with Playwright JS rendering, table extraction, and screenshots.

Tools (3)

Tool	Description
smart_scrape	Fetch any URL and extract content intelligently
screenshot_page	Screenshot URL → inline PNG in Claude
extract_tables	All HTML tables → clean JSON row arrays

Setup

cd /Users/dev/Desktop/development/azure/Fahad/web-lens
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python -m playwright install chromium

Claude Desktop Config

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "web-lens": {
      "command": "/Users/dev/Desktop/development/azure/Fahad/web-lens/.venv/bin/python",
      "args": ["/Users/dev/Desktop/development/azure/Fahad/web-lens/src/server.py"]
    }
  }
}

Test with MCP Inspector

cd /Users/dev/Desktop/development/azure/Fahad/web-lens
npx @modelcontextprotocol/inspector .venv/bin/python src/server.py