A pure-Python CLI that scrapes business listings (names, phones, addresses, ratings, tags…) into ML-ready DataFrames — and bypasses Cloudflare without launching a browser. The big rewrite: replaced the entire Selenium / undetected-chromedriver stack with TLS-fingerprint impersonation via curl_cffi. Same Chrome JA3/JA4 handshake on the wire, same client hints, zero chromedriver, zero 150 MB Chromium download.