wow-transmog-index

Build a reverse index of World of Warcraft item sources from Blizzard's official Game Data API — item → the boss that drops it, the profession that crafts it, the transmog set it belongs to.

Blizzard publishes the forward direction only. You can ask "what does Nefarian drop?" all day; there is no endpoint for "where does Bloodfang Hood come from?" This builds the other direction once, offline, so that lookup costs zero API calls at request time.

1,152 boss encounters across 211 instances
19,065 items mapped to their sources
 5,868 crafted items mapped to profession and expansion
 1,801 transmog sets mapped to 25,531 appearances

The whole index gzips to about 120 KB, which is small enough to ship inside a Cloudflare Worker bundle. That is the point: a Discord bot can answer "where does this drop" instantly, with no database, no cache, and no request-time failure mode.

Use it

You need a Blizzard API client — free, takes a minute.

export BLIZZARD_CLIENT_ID=...
export BLIZZARD_CLIENT_SECRET=...

node src/build-journal-index.mjs     # dungeon and raid loot  (~2 min)
node src/build-source-index.mjs      # professions and sets   (~10 min)

Both write ES modules into data/. Pass --dry to fetch and report without writing.

import { ENC, ITEMS } from "./data/wowmogdata.js";

// where does item 16908 come from?
for (const i of ITEMS["16908"] ?? []) {
  const [boss, instance] = ENC[i];
  console.log(`${boss} — ${instance}`);   // Nefarian — Blackwing Lair
}

Shape

Export	Meaning
ENC[i]	[bossName, instanceName]
ITEMS[itemId]	[encounterIndex, …] — 1,966 items drop from more than one boss
CRAFT[itemId]	[professionIndex, tierIndex]
APPSET[appearanceId]	setIndex
PROF TIER SETS	name tables the indexes point into

Names are stored once in a table and referenced by index. That is most of the difference between 120 KB and something you would not ship in a bundle.

What it does not cover, and why that matters

World drops, vendors, quest rewards, PvP and holiday events are not in here, because Blizzard does not publish them. The item endpoint has no source field at all. The dungeon journal covers dungeon and raid loot; professions cover crafting; everything else is genuinely absent from the API.

Wowhead has that data — from datamining plus user-submitted loot logs, rendered into HTML, with no API and terms that forbid scraping.

So a tool built on this index should say so when an item has no known source, rather than implying the item is unobtainable. An honest "this is not a dungeon or raid drop, here is the Wowhead page" is worth more than a confident wrong answer, and users notice the difference quickly.

Notes from building it

• Rebuilding costs ~1,260 API calls across 1,152 encounters, plus ~22,700 for the profession pass. Run it when a patch adds content, not on a schedule.
• A failed fetch must never read as "no source." Both builders retry, and refuse to write an index if any fetch failed — a silently incomplete index is worse than none, because nothing downstream can tell the difference.
• Sets repeat by name — Blizzard lists one per armour class, sometimes per gender. Collapse by name and keep the best match for the character you are answering for.
• Appearances are account-wide. A plate wearer really can be one piece from a cloth set. Filtering set progress by the character's armour type hides most of the answer.

Built for

This index powers a live Discord tool: type an item name, get the boss that drops it, the set it belongs to, and whether an easier item shares the same appearance. If you want that in your guild rather than building it yourself, Serverside builds guild tooling on Blizzard's API.

Changelog

See CHANGELOG.md. Current release: v1.0.0.

Licence

MIT. Built by Serverside — we build the software game communities run on.

World of Warcraft and the Battle.net API are trademarks of Blizzard Entertainment. This project is not affiliated with or endorsed by Blizzard.