# Upstream references

## Open Scriptures Hebrew Bible / morphhb

Repository: `openscriptures/morphhb`

Reference commit observed during v0.2 work:

`3d15126fb1ef74867fc1434be1942e837932691f`

The project imports XML from `wlc/*.xml`. The WLC text is identified by Open Scriptures as public domain; Open Scriptures markup is CC BY 4.0 with attribution.

## Open Scriptures Hebrew Lexicon

Repository: `openscriptures/HebrewLexicon`

Reference commit observed when building `data/lexicon/hebrew-lexicon.json`:

`21c9add13bc727d3a951361778e97e3ff7afd1ce`

## Septuagint — Swete

Repository: `nathans/lxx-swete` (derived from `OpenGreekAndLatin/First1KGreek`)

Reference commit used for `data/lxx.json`:

`26bad3eb42bba98471d154c954e36a6f30a0279d`

Also used for `data/interlinear/lxx-swete/`.

## Septuagint — Rahlfs ELS corpus

Repository: `alekkeersmaekers/glaux`

Reference commit used for `data/lxx-rahlfs.json`:

`b077d8f6ff429a5c7245954bc16bb7d1d7948823`

## Septuagint parallel — GLAUx

Repository: `alekkeersmaekers/glaux`

Reference commit used for `data/interlinear/lxx/`:

`b077d8f6ff429a5c7245954bc16bb7d1d7948823`

## Unicode/XML Leningrad Codex (UXLC)

Canonical site: `https://tanach.us/`

Mirror repository: `nerd-bible/tanach.us`

Reference commit used for `data/leningrad.json` and `data/leningrad-verses.json`:

`7b1d4e7a0776616eafa5a49c2b123acc87529e05`

Edition UXLC 2.4, build 27.5 (19 Oct 2025).

## Samaritan Pentateuch

Repository: `DT-UCPH/sp`

Reference commit used for `data/samaritan.json` and `data/samaritan-verses.json`:

`fb5b3db7fc5b42c6a5a33cc844664e0c054e8902`

Text-Fabric data version `7.2.4`.

## Russian Synodal with Strong's numbers — RST

Repository: `swmail/RST`

Reference commit used for `data/lexicon/ru-names.json`:

`78d6bbf343de07e06d07cf8bd32b7a38d85aea36`

## UBS open-license dictionaries

Repository: `ubsicap/ubs-open-license`

Reference commit used for `data/lexicon/ubs-*.json`:

`3a6edd8212df2e1189037ad39687726990c80d56`

## KARP Bible Code

Repository: `souldriver007/karp-bible-code`

Reference commit observed during v0.2 work:

`4650ad1c471feac5020c3b7c2de5482850e66363`

KARP is MIT licensed. Its repository contains ELS engine, proximity, statistics and UI modules. Hebrew Bible Code Lab is an independent Hebrew-first implementation and retains the KARP MIT notice as a reference attribution.

## Steinberg 1878 and Weisman 1899

Steinberg Unicode transcription: `attishn/tbnit` file `Tora.html`, used by `npm run build:historic-lexicons`.

Weisman OCR: Internet Archive identifier `Veysman_Grechesko-russkiy-slovar` (`Вейсман - Греческо-русский словарь_djvu.txt`).

Both source files are cached in `data/lexicon/source/` (gitignored). The derived JSON is committed.

## OpenAI Responses API (optional)

v0.5 uses an optional server-side HTTP call to `POST /v1/responses` only when `OPENAI_API_KEY` is set. The core ELS and Discovery computation does not depend on the API.