npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

magick-data

v0.2.1

Published

Magickal correspondences and reference data: the sephirot, planets, zodiac, Hebrew and Enochian letters, Golden Dawn grades, geomantic figures and more, as typed tables with the links between them

Readme

magick-data

Magickal correspondences and reference data — the sephirot, the planets, the zodiac, the Hebrew and Enochian letters, the Golden Dawn grades, the geomantic figures and the rest — as 29 tables of plain rows, a graph that says how they link, and the code that joins them. It is the data layer of magick.ly, which consumes it as anyone else would.

The tables are written as JSON5, which lives in the package's src/ in the repository and is not published, and each is built into a JavaScript module, a declaration and a JSON file. A row is a plain object; a link between tables is an id, and assemble() turns ids into the rows they name.

Licences

Copyright (c) 2020-2026 Gadi Cohen. The package is under two licences, by path, and LICENSE.txt says which file is which:

  • The code is MIT: the hand-written modules and what they compile to. Among them is astrology/kameas, which is code, not data: it computes the planetary squares, and writes out only the Sun's, a fact from Agrippa (Sources).
  • The data is CC BY 4.0: the JSON5 sources and everything the build emits from them, the Enochian Keys among them (Sources says, file by file, what is ours to license). Use it anywhere, including commercially, as long as you credit magick.ly.

The app that reads it is AGPL-3.0-or-later; this package is deliberately freer, so that it can travel.

The Enochian dictionary is neither. Its entries cite two works whose rights are not ours to give:

  • EMPM: Gerald J. Schueler, Enochian Magic: A Practical Manual (Llewellyn, 1985; second edition 1995), which is in copyright;
  • WE: The Whole Enochian Dictionary, an online compilation whose author and status are unknown.

So no rights are claimed in its entries, it is offered under neither licence, and it is not in the published package. It stays here for the app, at ./enochian/dictionary, which the published exports leave out.

Facts themselves belong to nobody. What is licensed here is the collecting, wording and arrangement.

Installing

npm install magick-data

It has no runtime dependencies, and it is ES modules only (Requirements). In Node, or anything that bundles:

import data from "magick-data";

const tiferet = data.sephirah.tiferet;
console.log(tiferet.name.roman, tiferet.planet?.symbol); // Tiferet ☉

TypeScript needs no setup: the declarations ship with the package, and every id is a literal type, so a misspelt one is an error. The three ways in below say what each kind of import costs, the tables list every subpath, and a reader that is not JavaScript can take the JSON. Wherever you use the data, credit magick.ly (Licences).

Three ways in

There are three ways to read the tables, and each costs something different. Two kinds of figure are given, and each says where it came from:

  • Type instantiations measured by the smoke test (pnpm --filter magick-data smoke, which prints them): the packed tarball, read by a consumer compiled with TypeScript 6.0.3 under bundler, with lib: ["es2022"] and skipLibCheck. Each of its reads loads the same 93 files, the lib's included.
  • Measured on the prototype at dc6f732: the bundle sizes, minified and gzipped by esbuild 0.28.2 (webpack 5.111.0 came within a few per cent), and the instantiations of a scoped assemble(), in a consumer compiled with TypeScript 6.0.3. These have not been measured on the shipped package.

The barrel, assembled, with every link resolved:

import data from "magick-data";

data.sephirah.tiferet.planet?.symbol; // "☉"

It holds 26 of the 29 tables: enochianTablet, christianChoir and seventyTwoAngel link to nothing and nothing links to them, so they are left out and read through their own subpaths. Reading one field costs 7,824 instantiations, and a four-hop read 17,621 (smoke test); the bundle is 17,290 bytes gzipped (dc6f732). Whatever imports it carries all 26 tables.

One raw table, by its subpath:

import sephirot, { type SephirahId } from "magick-data/kabbalah/sephirot";

sephirot.tiferet.planetId; // "sol", an id rather than a row

The rows are as authored, with no links. Reading a field costs no instantiations (smoke test), and sephirot is 1,650 bytes gzipped (dc6f732).

A scoped assemble(), over the tables you name:

import planet from "magick-data/astrology/planets";
import sephirah from "magick-data/kabbalah/sephirot";
import { assemble } from "magick-data/tools";

const scoped = assemble({ planet, sephirah });
scoped.sephirah.tiferet.planet?.symbol; // "☉"

Only the links between the named tables are made, and reading one that leads outside them is a type error. At dc6f732, two tables cost 5,637 instantiations, and sephirah and everything one hop from it, twelve tables, cost 11,588 instantiations and 8,838 bytes gzipped: more types than the barrel, because a partial scope's rows are a mapped type where the barrel's are named interfaces, and about half its bytes. A scoped assemble() is a bundle saving, not a type saving.

assemble() cannot be tree-shaken, since it is one call over the tables it is given; what bounds a bundle is which subpaths are imported. Calling it twice with the same tables returns the same object, and its rows are frozen.

The tables

Every table, by the name that assemble(), magick-data/tables and the graph use, with its rows and the subpath that exports it raw. A subpath is named after its file, so that it reads as the subject does; the table name is what a link's to names.

| Table | Rows | Subpath | | --- | ---: | --- | | planet | 15 | ./astrology/planets | | zodiac | 12 | ./astrology/zodiac | | astrologicalHouse | 12 | ./astrology/houses | | hebrewLetter | 27 | ./hebrewLetters | | enochianLetter | 21 | ./enochian/letters | | enochianTablet | 5 | ./enochian/tablets | | tetragram | 16 | ./geomancy/tetragrams | | geomanticHouse | 12 | ./geomancy/houses | | gdGrade | 12 | ./gd/grades | | gdDegree | 3 | ./gd/degrees | | archangel | 15 | ./kabbalah/archangels | | angelicOrder | 10 | ./kabbalah/angelicOrders | | christianChoir | 9 | ./kabbalah/christianChoirs | | fourWorlds | 4 | ./kabbalah/fourWorlds | | godName | 10 | ./kabbalah/godNames | | kerub | 4 | ./kabbalah/kerubim | | sephirah | 11 | ./kabbalah/sephirot | | treeOfLifePath | 24 | ./kabbalah/paths | | soul | 6 | ./kabbalah/souls | | tribeOfIsrael | 14 | ./kabbalah/tribesOfIsrael | | seventyTwoAngel | 72 | ./kabbalah/seventyTwoAngels | | chakra | 7 | ./chakras | | bodyPart | 12 | ./body/parts | | stone | 12 | ./materia/stones | | scent | 11 | ./materia/scents | | alchemySymbol | 10 | ./alchemy/symbols | | alchemyTerm | 6 | ./alchemy/terms | | element | 5 | ./alchemy/elements | | elemental | 4 | ./alchemy/elementals |

Most tables are objects keyed by id; astrologicalHouse, christianChoir and seventyTwoAngel are arrays. Each subpath's default export is the rows, and it exports the row's type and the table's by name (Sephirah, Sephirot). An object table's module exports its id's type as well, mostly as the row's name with Id (SephirahId). The exceptions are TetragramID, AlchemySymbolID, AlchemyTermID and EnochianTabletID; LetterId for enochianLetter; and planet, whose PlanetKey is every row's id while PlanetId names only the twelve whose kind is "planet". The three array tables' modules export no id type. magick-data/tables exports all 29 raw, with the Tables and TableName types, and magick-data/types and magick-data/rows the row types derived from the tables and the graph.

The graph

magick-data/graph declares every id-shaped field of every table, and magick-data/graph.json is the same for a reader that is not JavaScript. A field is one of:

  • links: it names a row of the table in to, or with many a list of them. mirrors names the field on the target row that must name this row back; both directions are stored, so that the JSON stands on its own, and the check holds the pair row by row. inverse derives an accessor on the target row pointing back here, which must be unique unless inverseMany makes it a list. as names the accessor where the field's name would give the wrong one.
  • pending: it names a table that does not exist yet, such as a planet's spiritId.
  • external: it names something outside the package, such as a tarot trump.
  • enum: its value is one of a fixed set, not a reference.

An accessor is named after its field: planetId gives planet, and planetIds gives planets, unless as says otherwise. A nested field is named by its dotted path, and its accessor lands at the same depth: treeOfLifePath's hermetic.hebrewLetterId gives hermetic.hebrewLetter. Optionality is never declared; a link a row does not have is an absent key or null.

Helpers

magick-data/tools exports all five, and assemble, rowOf and pathTarget have subpaths of their own as well:

  • assemble(tables) joins the named tables, as above (magick-data/assemble).
  • problemsOf(assembled) lists what the data says that the graph says it should not: a link that resolves to nothing, a mirror that does not answer, an inverse that is not unique, an accessor that would hide a field. assemble() never throws; it records problems instead.
  • accessorName(field, link) is the naming rule above, as a function.
  • rowOf(table, id) is one row by an id that is only a string, or undefined, so that an unknown id is a value the caller must handle (magick-data/rowOf).
  • pathTarget(table, path) says which table and field a dotted path such as gdGrade.planet.symbol ends at, by walking the graph, or undefined (magick-data/pathTarget). It imports every table, so it is for builds and tests rather than pages; a bundler drops it from a tools import that does not use it.

Texts and values that are not tables

  • magick-data/enochian/keys: the Enochian Keys, one row per word as Dee received it, with its English. A row holds each work's reading of the word (readings: Casaubon's print, with the word as dictated and his pronunciation column; the Golden Dawn's form in The Equinox I(8); room for Dee's own copy, Sloane MS 3191), where it is printed, and any note or doubt (uncertain). A version is a rule over the readings, as the tablets' are: keyVersion(key, version) gives each word in dee, the default (Dee's own copy when it can be read, else Casaubon's, else for a word his print lost Crowley's form, marked from), or gd, the Equinox's form, with a marker where Crowley joins a word to an earlier one, splits it, leaves it out or runs it into the word before. Both give Dee's English, and a word is always looked up by Dee's spelling (lookup). Key 19, the Call of the Thirty Aethyrs, marks the rows that change with the Aethyr as slots: aethyrCall(n, version) gives the Call for Aethyr n, and aethyrs holds the thirty with each work's spelling. Each Key has a checked record, the date, the works compared, the fields covered and what was not, and checkedText says it in a sentence, so that no page restates it. Where the Keys come from is in Sources.
  • magick-data/kabbalah/seventyTwoAngels/text: Lenain's entries on the seventy-two genii in full, in French and in English, loaded on demand.
  • magick-data/kabbalah/seventyTwoAngels/derived: what Lenain's four tables compute for each genius: degrees, days, minutes, sign, planet and choir.
  • magick-data/kabbalah/lenain/volume, …/notes, …/pieces and …/firstTable: Lenain's La Science Cabalistique leaf by leaf, the editorial apparatus, the text folded into paragraphs that run across leaves, and his first table reconstructed.
  • magick-data/astrology/kameas: the planetary kameas (magic squares), Saturn's 3 × 3 to the Moon's 9 × 9, keyed by the planet table's ids (kameas.luna) and frozen, row by row from the top. They are computed rather than written out, except the Sun's (Sources), so there is no JSON for them; oddKamea(n) and doublyEvenKamea(n) are the constructions, for an odd order from 3 or a multiple of 4, up to MAX_KAMEA_ORDER (100). Any other order is a RangeError.
  • magick-data/enochian/tablets: besides its rows, the tablets' versions. A row holds each source's reading of the tablet, square by square (readings: Regardie 1940, the Equinox 1912, Greer's edition, Dee's table of 1584 and Raphael's of 1587), and a version is a rule over them, not a copy: tabletVersion(tablet, version) gives its letters (null for an empty square) and the squares that hold more (alt, reversed, doubt), for gd, the Golden Dawn's and the default, regardie, equinox, dee and raphael (TABLET_VERSIONS), or undefined where a version has no such tablet: Dee and Raphael have no Tablet of Union. tabletDifferences(a, b) lists the squares where two versions differ, and how. A new source or version is a new key; no id moves.
  • magick-data/enochian/tabletNotes: notes on the Enochian tablets: each version, where it comes from and how it differs from each of the others (versionNote(id)), how the versions differ in general, each tablet, and particular squares, by row and column from 0 as the readings count them (squareNote(id, row, column)), with sourceText and sourceUrl for the works cited. The prose is ours, under CC BY 4.0 like the tables; the counts and squares where versions differ are the readings', never written into the prose.
  • magick-data/enochian/dictionary: for the app only, and not published (above).

JSON

For a reader that is not JavaScript, every table and text is plain JSON as well: magick-data/json/<path>.json, at its source's path (for example magick-data/json/kabbalah/sephirot.json), and the graph at magick-data/graph.json. The JSON holds the rows as authored. Nothing assembled is shipped, since an assembled result is cyclic and JSON cannot hold it.

Requirements

  • ES modules only. A CommonJS consumer under node16 resolution cannot import it; one under nodenext, and every bundler, can.
  • structuredClone, which assemble() uses: Node 17 and later, and every browser since 2022. The declarations do not mention it, so a consumer's lib needs nothing for it.
  • skipLibCheck: false costs about 64,000 instantiations, on a clean es2022 lib, whichever way in is used.
  • TypeScript 6.0.3 is what the package's smoke test checks a consumer with, under bundler and nodenext. The declarations carry .ts relative specifiers, which it resolves; a much older compiler may not.

Versions

Semantic versioning, applied to the data as well as the code:

  • renaming or removing a table, a field or a row's id is major;
  • new rows, fields or tables are minor;
  • a corrected value is a patch.

While the version is 0.x, a breaking change bumps the minor instead. The version says nothing about whether anything drawn from the data renders the same: a corrected value can change a picture, and a rename can leave every picture as it was.

From 0.1.0, each release is published by the repository's CI when the version changes on main, with npm provenance naming the commit it was built from. CI publishes a pre-release under the next tag and a release under latest. The first version, the release candidate 0.1.0-rc.0, was published by hand.

Romanised Hebrew

A romanised Hebrew name writes its prefixes as Israel's official romanisation does, the Academy of the Hebrew Language's system that BGN/PCGN adopted in 2018: the article, the conjunction and the prepositions written as prefixes in Hebrew are capitalised and joined to their word, "Roshit HaGilgulim", "Chayot HaKodesh", "YHVH Eloah VeDa'at". Where that word begins with an aleph, its vowel is capitalised in the same way, "Adonai HaAretz". The library standard, ALA-LC, writes prefixes in lower case with a hyphen instead; the data keeps to one of the two.

Sources

Most of this is correspondences, which belong to nobody. Every data file the package publishes is below, by its source in src/ (each is published as magick-data/json/<path>.json, beside its module and declaration), with where it comes from; the sections after the table name the works. A test holds the table to the files, so no data file ships without a row. "Signed off" is the date the owner checked the row's provenance, and no version past 0.0.0 ships while a row has none.

| Data file | Source | Signed off | | --- | --- | --- | | alchemy/elementals.json5 | Correspondences | 2026-11-12 | | alchemy/elements.json5 | Correspondences | 2026-11-12 | | alchemy/symbols.json5 | Correspondences | 2026-11-12 | | alchemy/terms.json5 | Correspondences; terms from the Second Knowledge Lecture, with no rights asserted over them | 2026-11-12 | | astrology/houses.json5 | Correspondences; interpretation is ours, following Lilly; motto.la is the mnemonic verse as usually quoted, not from Lilly, and motto.en our rendering of it (Lilly's houses) | 2026-11-12 | | astrology/planets.json5 | Correspondences; keyOfSolomon from the Key of Solomon; about and description are ours, written from public-domain works the file's header cites; intelligenceOfIntelligences, the Moon's, is the name and its Hebrew from Agrippa's table of the Moon, with 777's and Regardie's spellings of it beside his, names only, as the header cites them | 2026-11-14 | | astrology/zodiac.json5 | Correspondences; tetragrammatonPermutation is the Twelve Banners of the Name, Crowley's 777 (1909), col. CXL, as the file's header cites it: the twelve orders of YHVH's letters, one to a sign | 2026-11-14 | | body/parts.json5 | Correspondences | 2026-11-12 | | chakras.json5 | Correspondences of the seven-chakra system, facts only: names, places, petals, colours, seed syllables and their elements, and one-to-five-word glosses of the Sanskrit names. The Wikipedia "Chakra" article the file cites is a reference; no share-alike text is copied | 2026-11-12 | | enochian/keys.json5 | The Enochian Keys | 2026-11-12 | | enochian/letters.json5 | The Enochian letters | 2026-11-12 | | enochian/tabletNotes.json5 | Ours: notes on the Enochian tablets, citing their sources | 2026-11-14 | | enochian/tablets.json5 | The Enochian tablets | 2026-11-12 | | gd/degrees.json5 | Correspondences | 2026-11-12 | | gd/grades.json5 | Correspondences | 2026-11-12 | | geomancy/houses.json5 | Geomancy | 2026-11-12 | | geomancy/tetragrams.json5 | Geomancy | 2026-11-12 | | hebrewLetters.json5 | Correspondences | 2026-11-12 | | kabbalah/angelicOrders.json5 | Correspondences | 2026-11-12 | | kabbalah/archangels.json5 | Correspondences | 2026-11-12 | | kabbalah/christianChoirs.json5 | Lenain: the nine choirs of Pseudo-Dionysius, in his order and with his French | 2026-11-12 | | kabbalah/fourWorlds.json5 | Correspondences | 2026-11-12 | | kabbalah/godNames.json5 | Correspondences | 2026-11-12 | | kabbalah/kerubim.json5 | Correspondences | 2026-11-12 | | kabbalah/lenain/apparatus.json5 | Lenain | 2026-11-12 | | kabbalah/lenain/evidence.json5 | Lenain | 2026-11-12 | | kabbalah/lenain/pages.json5 | Lenain | 2026-11-12 | | kabbalah/paths.json5 | Correspondences | 2026-11-12 | | kabbalah/sephirot.json5 | Correspondences | 2026-11-12 | | kabbalah/seventyTwoAngels.json5 | Lenain | 2026-11-12 | | kabbalah/seventyTwoAngelsText/en.json5 | Lenain | 2026-11-12 | | kabbalah/seventyTwoAngelsText/fr.json5 | Lenain | 2026-11-12 | | kabbalah/souls.json5 | Correspondences | 2026-11-12 | | kabbalah/tribesOfIsrael.json5 | Correspondences | 2026-11-12 | | materia/scents.json5 | Correspondences | 2026-11-12 | | materia/stones.json5 | Correspondences | 2026-11-12 | | graph.json | Ours: how the tables link, built from src/graph.ts | 2026-11-12 |

Correspondences

The correspondences are traditional, the Golden Dawn's above all, and are facts: no one work is their source. What is ours is the collecting, wording and arrangement, which is what CC BY 4.0 covers. Where a file also holds text from an identified work, its row says so.

Lenain

seventyTwoAngels.json5, seventyTwoAngelsText/ and lenain/ come from Lazare Lenain, La Science Cabalistique (Amiens, 1823), read page by page from the Google Books scan in the repository's public/docs/. That scan is of a later reissue: it reproduces the 1823 title page and adds a preface by Papus for the Ordre Kabbalistique de la Rose-Croix. The volume does not date itself; the first reprint is recorded as Dujols and Thomas, 1909. Lenain died in 1877 and Papus in 1916, so no rights are asserted over either text.

What is offered under CC BY 4.0 is the work done on top: the English translated from that French, the choice and arrangement of fields, the editorial apparatus, and the corrections. No rights are claimed where none subsist. Nothing in it draws on any modern published translation of the work — the point of the exercise was a copyright-free rendering of the same public-domain material.

The Hebrew names are given both as Lenain points them and as bare letters. Forty-seven are there because two independent readings of the scan agree; twenty-five were read by a person who reads Hebrew, and each of those carries a note saying so. See plan 031.

The Key of Solomon

keyOfSolomon in astrology/planets.json5, on the seven classical planets, is what The Key of Solomon the King says each planet's days and hours serve for, in S. L. MacGregor Mathers' translation (London, 1888), Book I, chapter II, p. 12: his sentences verbatim, checked against sacred-texts' transcription of his edition. Mathers died in 1918, so the translation is in the public domain and no rights are asserted over the text. See plan 039.

Agrippa's kameas

astrology/kameas.ts, which is code rather than a data file, gives the squares of Heinrich Cornelius Agrippa, De occulta philosophia libri tres (Cologne, 1533), book II, chapter 22, which the Golden Dawn's diagrams follow. Agrippa died in 1535, the squares are facts, and no rights are asserted over them. Six are computed by the standard constructions, which give his squares, and the Sun's, of order 6, is written out. Two numbers of the 1533 printing are misprints, corrected to make the squares magic: Venus, row 4, column 5, printed 45 for the 43 of his Hebrew table, and the Moon, row 1, column 8, printed 45 for 54 in both his tables (archive.org DeOccultaPhilosophiaLoc1533, leaves n166 and n168). The tests check that every square is magic, that each starts with his first row (the Moon's as corrected), and that Saturn's and Jupiter's are his row by row. See plan 061.

The Enochian letters

enochian/letters.json5 gives each letter in both traditions, as the tablets do. english and the Golden Dawn's names (names.gd) are the Order's table in Israel Regardie, The Golden Dawn, vol. 4 (Aries Press, Chicago, 1940; archive.org IsraelRegardie-TheGoldenDawn-Vol4-1940, p. 299). Dee's names (names.dee) are those of the action of 26 March 1583 in British Library, Sloane MS 3188, whose f. 104 shows the letters as Dee drew them (Wikimedia Commons). sounds.gd is how the Order says a letter on its own, by the rule and examples of S. L. MacGregor Mathers and W. Wynn Westcott in the same volume (pp. 297 and 300), with Aleister Crowley's Golden Dawn reading of the Keys in The Equinox, vol. I no. 8 (London, 1912), where those pages give no example of the letter on its own. sounds.dee is the sound a letter stands for in Dee's spellings: his own notes beside the Keys, in Meric Casaubon's A True & Faithful Relation (London, 1659; archive.org truefaithfulrela00deej), with facts from Aaron Leitch, The Angelical Language, vol. II (Llewellyn, 2010), and Donald Laycock, The Complete Enochian Dictionary (1978), in our words. tetragramId is the geomantic figure the Golden Dawn gives sixteen of the letters, in Regardie, vol. 4, p. 77. The comment at the head of the repository's source file (letters.json5) cites the page or leaf of each; comments do not ship in the package.

Until 2026 the file also gave each letter a number (gematria), a tarot trump and a sign or element, all from Gerald J. Schueler, Enochian Magic: A Practical Manual (Llewellyn, 1985). They were removed: the numbers and trumps are his own derivation from the Hebrew letters, and neither Dee nor the Golden Dawn numbered the Enochian letters. The names, equivalents, sounds and figures are facts, and no rights are asserted over them; the wording and arrangement are ours, under CC BY 4.0.

The Enochian Keys

enochian/keys.json5 gives the Enochian Keys as Dee received them in Cracow from 13 April to 13 July 1584. Each Key is read word by word from the page images of Meric Casaubon's printing of Dee's journal, A True & Faithful Relation (London, 1659; archive.org truefaithfulrela00deej), which comes from British Library, Cotton MS Appendix XLVI: its Enochian with the pronunciation column beside it, and its English, word by word for Keys 1 to 4 and as running text for the rest, which we split into rows and say so. Each row and Key names its pages and leaves, and each Key's checked record says what was compared and what was not; Dee's own copy, Sloane MS 3191, whose images have been offline since 2023, is not.

None of the text the app had from 2023, whose source was not known, remains. Casaubon's print has lost the Enochian of parts of Keys 13 to 16 (damaged pages, pp. 133-138): two words of Key 13, most of Keys 14 and 15, and Key 16's third to fourteenth words. Those rows give Aleister Crowley's form from The Equinox, vol. I no. 8 (London, 1912), marked as his, until Dee's own copy can be read. The Golden Dawn's reading of each Key is Crowley's in the same Equinox, read word by word against Dee's rows. Key 19, the Call of the Thirty Aethyrs, is Casaubon's (pp. 201-209, the misbound leaves noted), with the thirty names as he prints them after it (p. 209) and as Crowley gives them (I(8) p. 125; I(5)'s chapter headings).

The words and their English are facts from public-domain prints, and no rights are asserted over them; the arrangement into rows, the split of the English and the notes are ours, under CC BY 4.0. Nothing in them comes from a modern edition or transcription, which are in copyright. See plan 073.

The Enochian tablets

enochian/tablets.json5 gives the Enochian tablets as each source prints them, so that both traditions are served. In the owner's words (24 October 2026): "we want to accommodate Enochian practitioners from both traditions; for gdHall we use the Regardie/GD version, but we also accommodate Dee purists." The Golden Dawn's version is the default. Its sources are Israel Regardie, The Golden Dawn, vol. 4 (Aries Press, Chicago, 1940; archive.org IsraelRegardie-TheGoldenDawn-Vol4-1940, pp. 270-275), The Equinox, vol. I no. 7 (London, 1912), Plates III and VIII (archive.org crowley-equinox_202110, the file "The Equinox Vol. 1, No. 7", leaves 241 and 246), and the redrawn colour plates of Regardie's book in its seventh edition, edited by John Michael Greer (Llewellyn, ISBN 978-0-7387-4399-8). Dee's table of 1584 and Raphael's reformed table of 1587 are read from Meric Casaubon's printing of Dee's records, A True & Faithful Relation (London, 1659; archive.org truefaithfulrela00deej, pp. 175-188 and the "Actio Tertia", p. 15), which comes from British Library, Cotton MS Appendix XLVI. Dee's own copy is expected in Sloane MS 3191, ff. 52r-80v, which the Library catalogues as the remainder of Liber Scientiae (catalogue); it could not be read: the Library's images have been offline since its cyber-attack of 2023 (plan 061 says what to check when they return). The comment at the head of the repository's source file (tablets.json5) says where each square comes from; comments do not ship in the package.

The greer reading is the letters as Greer's plates print them. Each is a letter that the 1940 or the 1912 print gives in that square, as its letter or as the smaller letter Regardie prints below it, except in one square, Water's row 4, square 9, where Greer gives e, the letter of Dee's 1584 table and Raphael's of 1587. Greer's case and his mirrored letters are typography. The Seniors' names, which settle some of the Golden Dawn's choices, are cited from Regardie, The Golden Dawn, vol. 3 (Aries Press, Chicago, 1939; archive.org IsraelRegardie-TheGoldenDawn-Vol3-1939, p. 73), which names all twenty-four. The letters are facts, and no rights are asserted over them, Greer's among them. See plan 061, "Tablet versions".

Geomancy

geomancy/tetragrams.json5: what each figure means in each house (meanings) is the Golden Dawn's table as Aleister Crowley printed it in "A Handbook of Geomancy", The Equinox, vol. I no. 2 (London, 1909), pp. 154-156 (archive.org crowley-equinox_202110, the file "The Equinox Vol. 1, No. 2", leaves n158-n160), read from the page images. The changes are editorial and listed cell by cell in plan 075: abbreviations spelled out, "evil" and "ill" read "unfavourable" (or "bad" where the Equinox pairs good with bad), and a full stop ending each entry. A figure's meaning is the Equinox's where it gives one, and otherwise our own words, from the Equinox's account or Agrippa's Fourth Book; translation is the plain English of the figure's Latin name.

geomancy/houses.json5 says what each house governs, in our own words, from Agrippa's list in "Of Geomancy", Henry Cornelius Agrippa his Fourth Book of Occult Philosophy, translated by Robert Turner (London, 1655), pp. 9-10 (archive.org b20458563, leaves n34-n35), with the matters Crowley ties to a house in the same Handbook (pp. 145, 153 and 159). The fifth and seventh houses follow Agrippa, not modern astrology.

Both works are in the public domain, and no rights are asserted over the Equinox's table or anything quoted from either; our own words and the edits are under CC BY 4.0. Nothing in them comes from a modern edition or essay. See plan 075.

Lilly's houses

interpretation in astrology/houses.json5 says what each of the twelve astrological houses signifies, in our own words, from William Lilly, Christian Astrology (London, 1647), Book I, chapter VII, "Of the twelve Houses, their Nature and signification", pp. 50-56 (archive.org b30338724, leaves n84-n90). The houses follow Lilly, not modern astrology: the father is the fourth house's and the mother the tenth's. Lilly died in 1681 and his book is in the public domain; our words are under CC BY 4.0, and nothing in them comes from a modern book or website.

motto is not from Lilly. Its Latin is the mnemonic verse as usually quoted, a word for each house: "Vita, lucrum, fratres, genitor, nati, valetudo, uxor, mors, pietas, regnum, benefactaque, carcer". We found no public-domain printing of it to cite. The table the file replaced had Iter for the ninth. The English renderings are ours.

The Second Knowledge Lecture

terms in alchemy/terms.json5 are the Golden Dawn's own short phrases for each alchemical term, from its Second Knowledge Lecture, one of the Zelator (1=10) grade papers, as Israel Regardie printed it in The Golden Dawn, vol. 1 (Aries Press, Chicago, 1937), pp. 119-120 (archive.org IsraelRegardie-thegoldenDawn-vol1-1937, leaves n110-n111), with small changes of form. They are the Order's wording of the 1880s and 1890s, not Regardie's, and each is a few words long. No rights are asserted over them, so the CC BY 4.0 grant does not cover them; the choice and arrangement of the table are ours.