Restored ~70 corrupted emojis across docs/index.html and website/index.html (matched contextually: nav, hero, channel-code, platform tags, feature icons, footer).
Added four-layer prevention so this doesn't ship again:
- UTF-8 BOM on the landing pages (first 3 bytes EF BB BF)
- .editorconfig pinning charset=utf-8-bom for docs/website index + support
- .gitattributes working-tree-encoding=UTF-8 for landing pages, CHANGELOG, README, GUIDE
- scripts/fix-emoji-encoding.py reusable repair script
Verified hashes match between docs and website copies. Google Ads tag AW-17950126697 already present (2 occurrences in each file).
