mycroft
por fabe
Search & ResearchResumen
EPUB and ebook ingestion, local vector index, and Q&A CLI for books.
Descripción
Qué hace esta skill
mycroft ingests EPUB files into a local vector index and lets you query book content via CLI. It supports semantic search, Q&A, and multi-turn chat sessions against any ingested ebook.
Storing embeddings locally means your book content stays private, unlike cloud-based reading tools that upload your documents.
Cuándo usarla
- Asking specific questions about a technical book without rereading it
- Searching for relevant passages across a long novel
- Running Q&A on a PDF-converted research book
- Building a personal library index of ebooks
- Resuming interrupted ingestion of large books overnight with batch mode
Flujo de trabajo de ejemplo
User asks: What does the book say about compound interest?
- 1
Run
mycroft book ingest /path/to/book.epubto index the EPUB - 2
Wait for ingestion to complete, or use
mycroft book ingest status <id>if batching - 3
Run
mycroft book search <id> "compound interest" --top-k 5to retrieve relevant passages - 4
Review the returned passages and synthesize an answer
- 5
Optionally run
mycroft book ask <id> "What does the book say about compound interest?"for a direct AI-generated answer
Quoted passages from the book with context, or a synthesized answer referencing specific sections
Requisitos
- OPENAI_API_KEY environment variable
Instalación
clawhub install fabe/mycroftEjecuta el comando de la CLI o pega el prompt en Claude Code, Cursor, Codex o el agente de programación que uses.