D&D Library Tools

Tools to work with D&D sourcebooks from the 5etools mirror as an AI reference library.

Setup: Sparse Clone

Clone only the book data (not the full website):

cd /path/to/your/libraries
git clone --filter=blob:none --sparse https://github.com/5etools-mirror-3/5etools-src.git 5etools-data
cd 5etools-data
git sparse-checkout set data/book data/books.json

To update later:

git pull

Structure

After cloning, you'll have:

5etools-data/
└── data/
    ├── books.json              # Index of all books
    └── book/
        ├── book-xphb.json      # Player's Handbook 2024
        ├── book-xdmg.json      # Dungeon Master's Guide 2024
        ├── book-xmm.json       # Monster Manual 2024
        ├── book-erlw.json      # Eberron: Rising from the Last War
        ├── book-efa.json       # Eberron: Forge of the Artificer
        ├── book-egw.json       # Explorer's Guide to Wildemount
        └── ...                 # 60+ books total

Book Abbreviations

ID	Name
xphb	Player's Handbook (2024)
xdmg	Dungeon Master's Guide (2024)
xmm	Monster Manual (2024)
phb	Player's Handbook (2014)
dmg	Dungeon Master's Guide (2014)
mm	Monster Manual (2014)
erlw	Eberron: Rising from the Last War
efa	Eberron: Forge of the Artificer
egw	Explorer's Guide to Wildemount
xge	Xanathar's Guide to Everything
tce	Tasha's Cauldron of Everything

Run node src/list.js to see all available books.

Tools

List Books

node src/list.js

Convert to Markdown (Optional)

node src/convert.js /path/to/5etools-data xphb erlw

JSON Format

The JSON is already excellent for LLM consumption:

{
  "data": [
    {
      "type": "section",
      "name": "Chapter Title",
      "page": 4,
      "entries": [
        "Plain text paragraph",
        {"type": "entries", "name": "Subsection", "entries": [...]}
      ]
    }
  ]
}

Tags like {@spell Fireball} reference other content and can be resolved or left as-is.