- Add rotating User-Agent pool (5 browser profiles) to all scrapers
- Fix Playwright: use browser context for UA instead of broken set_extra_http_headers
- Fix Selenium: set general.useragent.override preference
- Fix newspaper4k: pass browser_user_agent to bypass bot detection
- Add random delays (0.5-2s pre-fetch, 1-4s post-load) to mimic human behavior
- Switch Reuters RSS from Google News proxy to official Reuters agency feed
- Add legitimacy headers (Accept, Accept-Language, Connection) to Playwright
- Apply all fixes to both scraper.py and cron_scraper.py