One graph+vector record over humanity's sacred texts — and, in time, its deep history and its futures. When the sources disagree, it doesn't pick a side. It shows you both.
Read a question across every tradition at once — each stance grounded in a real, verbatim verse, with the disagreements shown, not smoothed over.
Convergence & divergence across the millennia — where traditions branch, borrow, and contradict, from the Enuma Elish to the living faiths.
Pure-stdlib, training-free, deterministic. It never generates prose — it quotes the corpus or refuses. Ask “How many gods are there?” and it will not answer as fact:
“Hear, O Israel: the LORD our God, the LORD is one.” — Deuteronomy 6:4
“He is Allah, the One.” — Qur'an 112:1
“To Indra and Agni and Varuna, the gods…” — Rig Veda
→ The corpus is the world's religions, and they disagree. Both citations stand. That refusal is the truest thing it can say.
The scripture corpus is the first layer. The project's aim is one continuous, cited record from the first origins to our possible futures — the same honest engine, more of the story:
Each new layer is tagged by stance — scripture, scholarship, or claim — and cross-linked, never flattened. It measures disagreement; it does not adjudicate truth.