About

This project sits inside the Orthodox tradition, not above it. It treats the corpus the way Orthodox theology has always read its fathers: with reverence for what they said, careful attention to where they differ from one another, and honest acknowledgment of where individual fathers were received critically by the conciliar tradition.

The corpus includes Origen of Alexandria, who is read with the caveat that the 553 anathemas condemn specific propositions attributed to him. It includes the Corpus Areopagiticum (Pseudo-Dionysius), which the modern Church reads as pseudonymous while continuing to cite as foundational for mystical theology. These inclusions are marked, not hidden, see the corpus list on the home page for the framing conventions.

The chat is a research assistant, not a teacher of the faith. It retrieves passages from the corpus and answers grounded in them, with citations. It does not speak for the fathers; it shows you what they wrote, and works hard not to invent quotations or flatten disagreements between them.

More than 110,000 paragraph-level passages from 280+ works by 78 authors, covering scripture (Septuagint OT, Greek Majority Text NT), the seven ecumenical councils and the locally accepted canons, the apostolic and ante-Nicene fathers, the post-Nicene Greek fathers (Athanasius, the Cappadocians, Chrysostom, Cyril of Alexandria, John of Damascus), Augustine and the Latin tradition, the Syriac fathers, the Corpus Areopagiticum, and supporting historical documents, every public-domain English translation we could responsibly source.

For the exact current count and a browsable index by section and author, see the library.

Material the project does not include, and the reasons:

• The Gnostic gospels (Gospel of Thomas, Mary, Philip, Judas, etc.), explicitly condemned by the Church as heretical. Including them without strong contextual framing risks misrepresenting Orthodox doctrine.
• Most modern English translations of the Eastern fathers after St John of Damascus , St Maximus the Confessor, St Symeon the New Theologian, St Gregory Palamas, St John Climacus, the Philokalia. The standard English editions are under copyright. The translation fund covers translating the underlying Greek public-domain text via Claude Opus, which is how the corpus would extend through the 14th century.
• St Photios's Mystagogy on the Holy Spirit, his anti-Filioque treatise, theologically vital but not available in any public-domain English translation we could find. The work waits on someone with both the rights and the philological care to do it well.
• Liturgical hymnography in translation: the Octoechos, Menaion, Triodion, and Pentecostarion are mostly under copyright in English, and faithful liturgical translation requires sensitivity to meter and tone that AI translation alone doesn't provide. Select feast hymnography may be added later through the translation fund; full liturgical books are out of scope for this project.

Each work is ingested from a public-domain English translation, parsed into paragraph-level passages with stable citations, indexed for keyword search (Meilisearch), embedded for semantic search (OpenAI text-embedding-3-small), and made available through the reader at /works/[slug]. The exact translation source and translator are recorded for every passage , visible in the reader and surfaced in the chat's citations.

The chat uses retrieval-augmented generation (RAG): when you ask a question, the system performs a hybrid keyword + semantic search across the corpus, picks the most relevant passages, and passes them to Claude Sonnet 4.6 along with strict instructions to ground every claim in the retrieved passages and cite them inline. The model is explicitly told not to invent citations, not to attribute quotes to passages that don't contain them, and to acknowledge disagreement among the fathers rather than presenting one view as universal.

A study tool, not a substitute for the Liturgy, the parish, the priest, or a spiritual father. Reading the fathers slowly and under direction is the traditional path; this project tries to lower the friction of finding what they wrote, not to replace the relationships within which Orthodox theology is meant to be received.

The AI chat is a research assistant. Its answers should be treated the way you'd treat a well-read seminarian summarizing a passage: helpful as a starting point, worth checking against the cited source, and never a substitute for prayer or conversation with a priest on matters of doctrine or spiritual life.

The corpus would not exist without the patient transcription work of others. With deep gratitude:

• Christian Classics Ethereal Library , the Ante-Nicene and Nicene/Post-Nicene Fathers series.
• Tertullian.org (Roger Pearse), Pusey's Cyril of Alexandria translations and Parker's Pseudo-Dionysius.
• Project Gutenberg , Mary H. Allies's 1898 translation of St John of Damascus's Three Treatises on the Divine Images.
• The World English Bible , public-domain English New Testament (Greek Majority Text).
• Sir Lancelot C. L. Brenton's translation of the Septuagint (1851), public-domain English Old Testament.

If you've spotted an error in a passage, a misattribution, or would like to suggest a public-domain source we should add, the project is grateful for the correction. (Contact channel TBD as the project moves toward public launch.)

The site is free to search and to read. The chat costs money to run because Anthropic charges per query, so it offers two paths: bring your own Anthropic API key (billed to your account, zero cost to the project), or become a $5/month Supporter and use the project's key (capped at ~$4 of compute per month). One-time donations toward translating specific gap-items are handled at /donate.

Memberships and donations are not tax-deductible. All payments are processed by Stripe; the project never sees card details.