About
Make the classical corpus
queryable, faithfully.
Why Living Śāstra exists.
The śāstras — the classical Sanskrit treatises on Ayurveda, Yoga, Vedānta, Tantra, and the rest — were designed to be conversed with. Generations of students sat with teachers and asked: What does this verse mean? How does this contradict the previous chapter? What does the tradition actually say about my situation?
That dialogic mode of study is hard to access now. Translations are scattered across out-of-print volumes, scholars are few, traditional teachers fewer still, and the texts themselves run thousands of pages each. The signal-to-effort ratio for an honest practitioner with a real question is brutal.
Living Śāstra exists to put the dialogic mode back within reach — by binding large language models tightly to curated public-domain editions of the canonical texts, and refusing to let them speak from anywhere else.
Our approach.
Citation-strict.
Every answer cites the specific passage it draws from. Click the citation to read the retrieved passage with chapter and verse. If supporting passages can't be found, the agent says so rather than invent.
Public-domain or cleared corpus only.
We index either editions clearly in the public domain (pre-1928 publications, translators long deceased) or works whose rights-holder has explicitly granted commercial deployment — no grey-area scraping, no licence violations, no royalty exposure.
No impersonation roleplay.
Agents don't speak as named teachers, gurus, or deities. They speak from the text. We treat the integrity of attribution as a load-bearing wall, not a feature toggle.
Refusing is a feature.
When a question falls outside an agent's corpus, when sources contradict without resolution, or when answering would require invention, the agent says so. Hallucinated certainty is the failure mode this product is built to avoid.
Not medical advice.
Ayurveda is a clinical tradition with thousands of years of refined practice. An AI conversation with the texts is a study aid, not a diagnosis. Don't substitute it for a licensed clinician. See our Terms.
What's next.
v1 ships three agents: the P.R. Sarkar (Shrii Shrii Ānandamūrti) discourse agent, covering sādhanā, social theory, and cosmology in English and Bengali; the Ayurveda agent (Caraka, Suśruta + Dutt’s Materia Medica); and the Yoga agent, spanning Patañjali’s Yogasūtras with the Vyāsa/Vācaspati commentaries, the Yoga-Vāsiṣṭha, the Bhagavad Gītā, and early practical manuals. The Haṭha Yoga Pradīpikā joins in a later release.
Future agents are gated on three things, in this order: a clean public-domain or licensed edition; a careful curation pass (chapter index, sense-map of terminology); and a small pilot with serious practitioners of that tradition before public release.
Candidate next agents include the Upaniṣads, the Bhagavad-gītā, the Yoga Vāsiṣṭha, and the principal Tantra texts in stable PD translation. If you are a scholar, teacher, or institution with a corpus to contribute — or with the rights to one we should licence — we want to hear from you.
Say hello.
Questions, collaborations, corpus contributions — all welcome.