Work Ingestion Pipeline

[Radiergummi]

Description

Complete ebook ingestion system with metadata extraction, duplicate detection, and entity creation.

Implemented

• ✅ Metadata extraction (EPUB/MOBI/PDF)
• ✅ Cover image processing with blurhash
• ✅ Asset deduplication by checksum
• ✅ Duplicate detection (checksum, ISBN, ASIN, fuzzy title)
• ✅ Creator/publisher normalization with fuzzy matching
• ✅ Series and tag extraction
• ✅ CLI commands (add, import, inspect)
• ✅ 119+ tests

Remaining Work

• Metadata enrichment integration (wire up providers)
• Enrichment UI (review/accept metadata)
• External identifier linking (VIAF, ISNI)
• Cover fetching from OpenLibrary
• Language detection
• Calibre library import

Files

• Plan: plans/work-ingestion.md
• Ingestion: packages/sdk/src/ingestion/
• CLI: apps/cli/src/commands/works/

Getting Started

1. Review existing ingestion pipeline
2. Wire up metadata enrichment
3. Build enrichment review UI