- Add layer_count and multilingual layer_names (id, number, slug, name, description) to narrators_json in hadiths SQLite table
- Include order, chain_index, is_gap, layer/layer_slug/layer_level/layer_name, and full identity fields (arabic_full_name, known_as, arabic_known_as, nickname, arabic_nickname, kunya, arabic_kunya) on transmitter nodes and narrators table
- Return all chains' transmitters when ?layer query param is omitted on GET /api/hadis/<slug>/transmitters/ and always return complete layer_names list
- Add source_type, sect_type, and categories_json columns to hadiths table and source_type/sect_type to hadith_categories
- Resolve category sect_type from HadisCategory.sect relation and inherit source_type/sect_type from ancestor categories when unset
- Add reference filter option tables (sources, authors, narrators, statuses, tags) with indexes for 100% offline All Arguments filtering
- Include book cover image URL in serialized reference book objects and full narrator/uncertain_transmitters metadata in narrators_json
- Add links_json column to hadiths table and standardize link objects via format_links_output across hadiths, interpretations, and corrections
- Include complete book, author, book_authors, edition/edition_info (title, language, edition_number, publisher, publication_place, contributors with role order and volume scope, researchers, editors), and volume/citation fields in references_json for all exported tables
- Optimize HadisSyncSerializer: remove unused number, address_details, and book descriptions
- Separate sync serializers for corrections and interpretations (omitting redundant hadis_info and category_info)
- Add HadisDatabaseVersion model and migration for tracking offline SQLite DB
- Add export_hadis_sqlite management command with gzip compression and single-file overwrite
- Add GET /api/hadis/sync/database/ endpoint and enhance version serializers
- Automatically trigger background SQLite database export on version increment in admin