Simplify Miruro: direct M3U8 scraping from page

Improve Miruro extraction: wait longer for video + SSR monkey decode

Fix Miruro endpoint: remove server-side CompressionStream

Add /api/miruro endpoint for browser-context pipe extraction

Update debug to show page structure/buttons

Block devtools-detector + recaptcha scripts, wait longer for SPA

Add /api/debug endpoint for diagnosing extraction issues

Fix Dockerfile: use node:22-bookworm, install Chromium via Playwright

Add README with Docker config

Initial commit: Playwright stealth proxy with Express API