Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Ogrods/BAKLOG/llms.txt

Use this file to discover all available pages before exploring further.

Baklog’s enrichment pipeline has three layers: behaviours that run automatically while the app is open, scheduled scripts you can configure to run while the app is closed, and individual enricher scripts that backfill specific metadata columns. Together they keep your library’s hours, reviews, covers, deal prices, and compatibility tiers current without requiring constant manual intervention.

​
While the app is open

Three automatic behaviours run in the background whenever python server.py is active and you have the dashboard open. All three are on by default and can be toggled from the Connections tab. Auto-refresh stale stores quietly refreshes one store older than 24 hours every ~30 minutes. Only one store refreshes per cycle so the process stays unobtrusive. Toggle it off on the Connections tab if you prefer manual control. ITAD auto-refresh pulls fresh deal prices from IsThereAnyDeal on a 15–60 minute schedule from the Fetcher health panel. Prices in the Wishlist tab stay nearly current without any manual action. Auto-enrich watches for library fetches that add new games and then automatically queues the HLTB, reviews, covers, and co-op tag enrichers for those new entries. Metadata backfilling happens in the background while you browse.
On the free tier, store refresh runs one store at a time, on demand. The paid tier ($5/mo) adds scheduled refresh that runs even when the app is closed — no scripts required.

​
Fetcher health panel

The Fetcher health row in the dashboard is the live control panel for all fetchers and enrichers.
  • Click any chip to enqueue that fetcher. Output streams live into the log panel and the chip refreshes when the run finishes.
  • Shift+click on a library or wishlist chip adds --refresh (ignores cache, refetches everything).
  • Shift+click on HLTB, Reviews, or Covers enricher chips adds --retry-misses (re-attempts rows previously cached as no-match).
Stall watchdog — when a fetcher runs via server.py and stdout is silent for 30 seconds, the server injects a [server] no output for Ns - still running (PID …) message into the log panel, repeating every 60 seconds. This is informational only — the process is never killed automatically. Run logs land in profiles/<id>/cache/runs/*.jsonl for later inspection.

​
Scheduled refresh (app closed)

For refreshes while Baklog is closed, use the OS-appropriate helper script. The Pro tier adds this capability through the tray or OS scheduler without any manual script setup.
Run the PowerShell helper from your Baklog directory:
Set-Location "C:\path\to\steam-backlog"
.\refresh.ps1
To refresh on a schedule without keeping the app open, register a weekly scheduled task. This example runs every Sunday at 9:00 AM:
schtasks /create /SC WEEKLY /D SUN /TN "BAKLOG Refresh" /TR "powershell -ExecutionPolicy Bypass -File \"C:\path\to\steam-backlog\refresh.ps1\"" /ST 09:00
Scripts and UI runs use the same fetch sequence and log to refresh.log.

​
Enrichment scripts

Enrichers backfill specific metadata columns on existing library rows. Run them directly or via the python -m enrichers wrapper.
ScriptPurposeCommand
enrich_steam_reviews.pyBackfill Steam review % on non-Steam rows via Steam Store searchpython enrich_steam_reviews.py
enrich_cross_store_images.pyBackfill header_image / library_image from Steam CDN for non-Steam rowspython enrich_cross_store_images.py
enrich_hltb.pyBackfill HowLongToBeat hours on any games_*.json row missing thempython enrich_hltb.py
enrich_protondb.pyBackfill ProtonDB Linux / Steam Deck compatibility tiers on Steam-matched rows; no API key requiredpython enrich_protondb.py
fetch_itad.pyPull cross-store deal prices into itad_prices.json; also refreshes FX ratespython fetch_itad.py
fetch_fx.pyRefresh FX rates only (cache/fx_rates.json via Frankfurter, 24h cache)python fetch_fx.py
Enrichers wrapper — the python -m enrichers command provides a convenient entry point for the three most common enrichers: 
python -m enrichers hltb
python -m enrichers steam-reviews
python -m enrichers cross-store-images

​
Exit codes

Every fetcher script prints a === name started at … === header and a footer summary with elapsed time. The exit code tells you exactly what happened:
CodeMeaning
0Success
1Runtime or configuration error
2Suspicious empty result (or ITAD resolved zero titles) — stale data is preserved by default
3Drift guard refused to write (result deviated too far from the cached version)
4Auth failure — credential is expired or invalid
Fetcher failures appear in the Fetcher health panel and in run logs. They are not automatically sent to any bug-report endpoint. See Getting help if you need to file a report.

​
Fetcher flags reference

These flags work across most fetcher and enricher scripts. Pass them on the command line or trigger them via Shift+click in the Fetcher health panel.
FlagEffect
--refreshIgnore cache and refetch everything from the source
--retry-missesRe-attempt enricher rows previously cached as no match
--only-newOnly fetch games not already present in the store JSON file
--skip-hltbSkip HowLongToBeat lookups (faster run, leaves existing HLTB data untouched)
--allow-emptyAllow writing a zero-item result (bypasses the suspicious-empty guard)
For per-store connect steps and CLI fallbacks, see Connecting stores.

​
Data attribution

Baklog surfaces third-party metadata from the following providers: Store logos and trademarks belong to their respective owners. Baklog is not affiliated with Valve, Epic, GOG, or other storefronts.

Build docs developers (and LLMs) love

Get started for freeTalk to us