I need a lightweight Node.js app that crawls a handful of public web pages, parses the information I specify, and forwards the fresh data to my Telegram bot every 2-5 minutes. The script should run unattended on a Windows machine, recover gracefully from connection hiccups, and make it easy for me to add or remove target URLs without touching the core logic. Core flow 1. Fetch pages (headless if necessary). 2. Extract the required public data with cheerio, puppeteer or a similar tool. 3. Push neatly formatted messages through node-telegram-bot-api. 4. Repeat on an adjustable interval, defaulting to 2–5 minutes. Acceptance criteria • Clean, well-commented Node.js source (ES6+) delivered in a single Git repo. • A simple config file for URLs, selectors, Telegram token, chat ID, and polling interval. • README with setup steps and a one-command start script. • Demonstration of the bot posting live data on my test channel. Speed is important—I’d like the first working version as soon as possible. When you reply, highlight your relevant experience building scrapers, working with Telegram bots, and handling scheduled jobs in Node.js.