Integrations¶

These integrations are specific to the fork. For Plex, Radarr, Sonarr, and TMDB setup, see the upstream docs.

Cloudflare Solver¶

Networks Top 10 collections pull data from FlixPatrol, which sits behind Cloudflare. They need a solver. The image includes a headless Chromium browser that Agregarr uses when you haven't configured one, but it can't reliably get past FlixPatrol's challenge. If you have a Networks collection, the Cloudflare Solver health check reports an error until you add a solver.

Supported solvers¶

• FlareSolverr
• Byparr
• Trawl

Setup¶

1. Run a solver next to Agregarr. There's a compose example below.
2. In Settings > Sources, find Cloudflare Solvers under Fetching Settings. Press Add Solver, give it a name, and enter its URL, for example http://flaresolverr:8191.
3. Add more instances if you want a fallback. Agregarr tries them in order. Drag a row using Drag to reorder to change its priority.
4. Press Save Changes.

services:
  agregarr:
    image: bitr8/agregarr:latest
    # ... your existing config

  flaresolverr:
    image: ghcr.io/flaresolverr/flaresolverr:latest
    container_name: flaresolverr
    environment:
      - LOG_LEVEL=info
    ports:
      - "8191:8191"
    restart: unless-stopped

Use :latest for the Agregarr image, or pin a release such as :2.13.0.

When both containers share a network, use the service name. Otherwise, use the host IP and published port.

Behaviour¶

• Once you've configured a solver, Agregarr won't fall back to the built-in browser. If every solver is down, the health check reports it.
• FlixPatrol always goes through the solvers, or through the built-in browser when you haven't set one. Plain HTTP isn't an option for it.
• The first failure from an instance is free. After two failures in a row, Agregarr skips that instance for that site for 1 minute. Each failure after that doubles the wait, up to 16 minutes. If the next failure happens more than 20 minutes after the previous wait ended, the count starts again. Other instances keep going.
• Fetched pages stay cached for 5 minutes.
• If FlixPatrol's overview page doesn't have a section for a platform, Agregarr reads that platform's own Top 10 page instead. Peacock and Hulu use FlixPatrol's United States page.
• Sometimes a solver loads the page and the site still refuses it. The log says so, for example solver loaded flixpatrol.com but the site returned HTTP 403. If every solver is refused, the health check stays red for up to 24 hours, or until a later fetch works. Cloudflare may be challenging your server's IP address. Run the solver through a VPN.
• The old single "FlareSolverr URL" setting moves to this list automatically.

Letterboxd¶

Letterboxd normally works over plain HTTP, but some VPS and VPN IP ranges get Cloudflare challenges. Turn on Use Cloudflare Solver for Letterboxd in Settings > Sources to route requests through your solvers. If you haven't configured a solver, it falls back to the built-in browser. Leave this off unless Letterboxd is actually blocking you. Plain HTTP is faster. While the toggle is off, a health check warns if Letterboxd was blocked during the last 8 hours.

Tracearr¶

You can use Tracearr instead of Tautulli for watch statistics. Its settings are under Settings > Sources, in Tracearr Settings.

1. Enter Hostname or IP Address, Port (default 3000), Use SSL, URL Base if you have one, and your API Key.
2. Press Test Connection. When it connects, the Server list fills in. Leave it on All servers to count plays from every server Tracearr tracks, or choose one.
3. Turn on Use Tracearr for statistics and press Save Changes.

While it's on and both the host and API key are set:

• The statistics collection type is labelled Tracearr Statistics instead of Tautulli Statistics.
• Collection statistics and weekly play counts on the dashboard come from Tracearr.
• Agregarr asks Tracearr for no more than the last 365 days. Longer ranges are capped at 365, and the dashboard Days: box only accepts 1 to 365.
• Results stay cached for 5 minutes.

If Agregarr can't read from Tracearr, the dashboard shows Tracearr Connection Failed with a Check Tracearr Settings button. Turn the toggle off to switch back to Tautulli.

Seerr¶

Set up Seerr under Settings > Downloads, in Seerr Settings.

The Keep placeholder ignore patterns in Seerr option is in the Seerr connection window. It's off by default. When you turn it on, Agregarr adds the two patterns matching its placeholder files to Seerr's Ignored Path Patterns. That stops Seerr counting placeholders as available media. Agregarr checks them during every collection sync and puts back any pattern that's gone missing.

The window shows the current state beside the checkbox: Both patterns present, {count} of 2 patterns missing, or Seerr build doesn't support Ignored Path Patterns.

This needs a Seerr build with an Ignored Path Patterns setting. The checkbox help text lists rubeanie/seerr:ignore-media-regex as one that has it. The Seerr Placeholder Patterns health check reports any problems.

Maintainerr¶

Connect Maintainerr to put deletion countdowns on movie, show, and season posters. Maintainerr's collection schedules provide the dates.

You'll find the connection fields under Settings > Sources > Maintainerr Settings (Poster Overlays Only): hostname, port (default 6246), API key, and SSL. Test Connection checks the details before saving. A health check catches problems later.

Days remaining is deleteAfterDays minus the number of days since the item's addDate. If an item belongs to more than one collection, the soonest date wins.

Season countdowns need Maintainerr 3.4.0+. Older versions return the collection type as a number, so Agregarr ignores those collections. Movie and show countdowns work with any version.

Full details: Maintainerr Countdown.

Sonarr-First Next-Episode Dates¶

The fork asks Sonarr first for a show's next episode date. If Sonarr doesn't track the show, it falls back to TMDB. A failed Sonarr lookup means "unknown", not "no upcoming episode", so an outage doesn't wipe existing countdowns.

There's nothing else to configure. If Sonarr is connected, Agregarr checks it first.

IMDb charts¶

The IMDb Top 250, Top 250 English, Bottom 100, Popular, and Box Office collections read from IMDb's public GraphQL API. Agregarr rejects responses with fewer items than expected for every chart except Box Office.

TMDB Trailer Resolution¶

Before searching YouTube, the fork checks TMDB's video metadata for a placeholder trailer. The lookup runs in this order:

1. TMDB videos. Only YouTube trailers and teasers count. Trailers come before teasers, official uploads before fan uploads, and newer videos first. Agregarr tries up to three and accepts videos under 10 minutes.
2. YouTube search. Agregarr searches for the title, year, and "official trailer", then checks the top 10 results. It downloads the first result that passes the word lists if it's under 3 minutes 30 seconds. If it's longer, Agregarr uses the static placeholder video.
3. Static placeholder video. Used when neither source finds a trailer.

The same downloader that fetches trailers also runs the YouTube search. It uses your youtube-cookies.txt file if you've added one.

Settings¶

All of these are under Settings > Downloads:

Downloads are capped at 1080p.

YouTube cookies¶

Export cookies from a logged-in browser in Netscape format, then save them as youtube-cookies.txt in your Agregarr config folder. Settings > Downloads shows whether Agregarr found the file. If it's empty or isn't in Netscape format, the screen shows YouTube cookies file is not valid and downloads continue without it. The collection form shows a YouTube Cookies Not Configured warning when the file is missing or invalid.