Answer book questions with source-backed citations from a QMD collection.
npx skills add https://github.com/0xtito/second-brain --skill qmd-book-qa
Please help me install this Agent Skill. Skill: qmd-book-qa Source: https://github.com/0xtito/second-brain/tree/main/pi-extensions/skills/qmd-book-qa Command: npx skills add https://github.com/0xtito/second-brain --skill qmd-book-qa
Enables precise, evidence-backed answers to user questions about a single book indexed in QMD, pulling citations only from that book's collection.
Ask a specific question about a book and provide an optional page hint to retrieve grounded citations.
High-intent search queries and answers about installing and using this skill.
To get grounded answers with exact book sources, you query a specific book indexed in QMD. It operates within a single book collection at ~/books-md and returns targeted citations referencing exact docids and file paths.
Yes, you can provide optional page-level or chapter-level hints when querying book content. This scoped querying targets exact content within the indexed collection to retrieve highly specific, source-backed citations.
Supported retrieval modes for searching a single book include search, vsearch, and query. These modes support explicit docid-based hydration to ensure precise, evidence-backed results pulled directly from the indexed collection.
Yes, you need a valid QMD collection to extract citations from a book. The system requires a pre-indexed collection located at ~/books-md and uses a bundled helper script to resolve targets and generate ready-to-run QMD get commands.
A key limitation of single book evidence retrieval is that it operates strictly within one book's collection at ~/books-md. It cannot pull citations or cross-reference evidence from multiple books or external documents outside that specific collection.
johncarpenterSearch, retrieve, and index local Markdown files using the QMD CLI.
cdeistopenedIndex and search local Markdown collections with BM25 and vector embeddings.
w-winterSearch indexed text corpora with semantic ranking and structured queries.
JCETools-PetraDesign RAG pipelines with chunking, embedding, retrieval, and evaluation metrics.
