steam-appdb

Open database of all Steam app ids — games, dedicated-server tools, DLC, software, music, videos, hardware — split into per-type JSON files. Built over anonymous PICS (the keyless Steam-client protocol steamcmd itself uses): no account, no API key.

Why

Valve removed the keyless ISteamApps/GetAppList, and its key-gated replacement exposes no tools — so dedicated servers can't be listed through the Web API at all. Anonymous PICS still sees everything; this repo keeps that data public and fresh.

How it updates

.github/workflows/update.yml runs the pipeline every 2 hours (cron 17 */2 * * *, plus manual workflow_dispatch):

1. checkout + pip install -r requirements.txt;
2. python dump_applist.py --data-dir data — anonymous PICS login, asks the changes feed what changed since the changenumber stored in data/meta.json, re-probes only those app ids (seconds per run), rewrites the files in data/;
3. commits and pushes data/ if anything changed.

New releases emit PICS changes, so they arrive automatically; the id-space scan below is only ever needed once. Manual run with a scan_to input does a backfill scan instead.

Repository layout

dump_applist.py               the dumper (anonymous PICS, incremental)
verify.py                     sanity checks for a bootstrapped data/
requirements.txt
.github/workflows/update.yml  the update pipeline (every 7 hours)
data/                         generated:
  all.json                    [{appid, name, type}] — master list
  game.json, tool.json, dlc.json, application.json,
  music.json, video.json, demo.json, hardware.json, ...
                              [{appid, name}] per product type
  applist.json                {"applist":{"apps":[...]}} — the
                              steamcmd-installable subset (game+tool+application)
  meta.json                   PICS changenumber + scanned ceiling

Setup

1. Create a public repo and copy this directory's contents into it.

2. Bootstrap the database once (locally is fine, takes roughly 30–60 minutes — batches of 5000 ids per PICS call, checkpoint every 100k):

   pip install -r requirements.txt
   python dump_applist.py --scan-to auto
   git add data/ && git commit -m "bootstrap" && git push

   auto scans the id space from the beginning and stops by itself once 200k consecutive ids turn out empty — i.e. past the end of Steam's assigned id range (an explicit integer ceiling is also accepted). If the run dies, re-run the same command: it resumes from the last checkpoint. Ids created later arrive through the changes feed, so the scan is never repeated.

   Check the result before pushing: python verify.py (file consistency + spot checks of well-known apps and dedicated servers) — expect ALL GOOD.

3. Done — the pipeline keeps data/ fresh from here on.

Wiring into gamefetcher

https://github.com/Austrum-lab/game-fetcher-cli

Point at the installable subset (single URL, classic GetAppList shape), or at individual per-type files — both shapes are parsed:

# gamefetcher.yaml
app_list_urls:
  - https://raw.githubusercontent.com/<owner>/steam-appdb/master/data/applist.json

Once this repo is live it can replace all of gamefetcher's default sources.