Scrobble
Internal blueprint for the scrobble pipeline (event parsing, filtering, and sink dispatch).
Last updated
Was this helpful?
⌘Ctrlk
Internal blueprint for the scrobble pipeline (event parsing, filtering, and sink dispatch).
This section documents the scrobble pipeline under providers/scrobble/.
It also covers the legacy webhook bridge under providers/webhooks/.
Scrobbling pushes playback progress to trackers in real time.
It uses three pieces:
providers (Plex/Jellyfin/Emby) that produce events
a dispatcher that filters and dedupes
sinks (Trakt/SIMKL/MDBList) that send events outbound
core model + dispatcher: providers/scrobble/scrobble.py
watch providers: providers/scrobble/<provider>/watch.py
sinks: providers/scrobble/<sink>/sink.py
runtime state writers: providers/scrobble/currently_watching.py, providers/scrobble/_auto_remove_watchlist.py
legacy webhook bridge: providers/webhooks/*.py
ScrobbleEventCommon fields:
action: start, pause, stop
media_type: movie, episode
ids: IMDb/TMDb/TVDb tokens (when present)
progress: 0–100
server_uuid, username, session_key for filtering and dedupe
raw payload for debugging
Supported Plex shapes:
PlaySessionStateNotification (from_plex_pssn)
flat playing objects (from_plex_flat_playing)
webhook wrapper (from_plex_webhook)
ID extraction relies on Plex GUID patterns:
imdb://tt...
tmdb://...
thetvdb://...
The dispatcher enforces:
server UUID + username allowlists
pause debounce (defaults to 5s)
per-session last action and last progress tracking
/config/.cw_state/currently_watching.json
/config/.cw_state/watchlist_wl_autoremove.json (TTL dedupe)
Last updated
Was this helpful?
Was this helpful?