A Citation-First Research Index
The Mishnah Oracle
Ask a question; the Oracle answers from the DNCZ corpus — about 8.17 million file-atoms indexed with SQLite FTS5 and searched at C-speed from inside davar_http. Every answer is grounded in cited atom IDs you can verify. No external model, no scripts.
The simple believeth every word: but the prudent man looketh well to his going. — Proverbs 14:15
Ask the Oracle
Type a question above. The Oracle retrieves the most relevant atoms from the live corpus index and answers with citations — about 8.17 million file-atoms by name and path.
How the Oracle Answers
Your question runs through DAVAR's C handler (handler_dncz_find) as a ranked full-text MATCH over the DNCZ corpus index — a native SQLite FTS5 database of about 8.17 million file-atoms. The Oracle takes the top-ranked atoms, frames what the corpus surfaced, and renders the result here as plain HTML with each supporting atom cited by ID and path. There is no client-side JavaScript and no Python at serve time, and no answer is generated from outside the retrieved corpus.
This is a citation-first research index over estate metadata — atom names and paths — not an infallible proof engine and not a doctrine authority. For Scripture-first study and confidence-labeled evidence, see /bible-search, /apologetics, and /research-articles.
Verify and Go Deeper
The Oracle points you at evidence; it does not replace it:
/dncz-search — search the same corpus directly by term.
/dncz-proof — what DNCZ is, with honest claim boundaries.
/bible-search — search all KJV verses.
/research-articles — evidence-lane summaries (confidence labeled).
/apologetics — five-domain overview.