OpenAI Whisper chunking, article cleaner, video transcribe, re-sync safety

Article quality harness, attach article, PDF-ready refactor prep

Phase 3: Hybrid Video Transcription implemented

Fix: media download extensions and article extraction logic verified

Allow Tampermonkey sync on any X page, not just bookmarks

Fix: articles now extract for re-synced bookmarks (UPSERT), add local image download

Fix: Tampermonkey script broken backtick escaping - use string array instead

Fix: auto-scroll, text extraction, link capture + add test harness

Phase 2: article extraction, auto-scroll, export, bulk delete, enhanced UI

merge remote, keep local README

first commit

Initial commit