- Update getCurrentMediaData to parse script tag with id 'story-data'
- Extract data from attributes.pictures for pictures:
  * title, caption, description, tags_names as tags, user_tags_ids as userTags
- Extract data from attributes.videos for videos:
  * title, description, tags_names as tags, user_tags_ids as userTags (no caption)
- Add fallback to direct attributes and root level data
- Update both content.js and background.js injected functions
- Preserve existing tags and user tags when updating privacy
- Use parsed data in XHR payloads instead of empty defaults