SparkBox patches — Sat Oct 3, 2026
Posted by Chris
SparkBox releases shipped on Sat Oct 3, 2026 (UTC), in order, newest at the bottom. This thread is updated as releases land during the day. Update from the SparkBox dashboard, or run sudo sparkbox upgrade on your SparkBox host. Legend boxes get every release the day it ships; free boxes get it after it has run there for a week. If a problem remains after updating, submit a private support ticket or reply below with the app name, what you expected, and what happened instead. Earlier days: Fri Oct 2, 2026 · Thu Oct 1, 2026 · Wed Sep 30, 2026 · Tue Sep 29, 2026 · Mon Sep 28, 2026 · Sun Sep 27, 2026 · Sat Sep 26, 2026 v1.6.840 — Plex can be linked to your Plex account again when it says "Server already claimed" - If your Plex server was removed from your Plex account (or linked to the wrong one), Plex kept the old link saved on your box and quietly ignored every new claim code, so the only way out was editing Plex's settings file by hand. Now get a fresh code at plex.tv/claim while signed in to the account that should own the server, and within 4 minutes run: sudo sparkbox plex-reclaim claim-xxxxxxxx. It removes only the old link (keeping a backup of Plex's settings), applies your code and tells you whether Plex accepted it. Your libraries, watch history and settings stay as they are. v1.6.841 — Frigate's admin password can be found again when the dashboard says it hasn't been generated - Opening Frigate from the dashboard could show no password, only "Credentials haven't been generated yet — the Media module is still finishing setup", while Frigate's own login page asked for one. Frigate shows its first password only once, the first time it starts, so after an update or after turning it on again the dashboard had nothing left to show. Now run: sudo sparkbox frigate-password. It gives Frigate a new admin password, prints it, and saves it so the dashboard shows it when you open Frigate and under Settings → Passwords. Your cameras, recordings and other Frigate users are not touched. The window that opens with Frigate now says this instead of mentioning the Media module. v1.6.842 — Tom AI can free the space Docker reports as reclaimable - When the disk was filling up and SparkBox reported several GB of reclaimable app images, asking Tom AI to free space got "there is no cleanup button", even though Settings → Tools → Run cleanup frees exactly that space. Tom AI now offers that same cleanup: it says how much space it expects to free and what goes (older versions of your apps kept from past updates, after which Roll Back to those versions is no longer possible, and the images of apps you switched off, which download again if you switch them back on), and it runs only after you confirm with your dashboard password. Your files, settings and running apps are not touched. Unused Docker volumes are explained but never offered for removal, because a switched-off app can keep its data there. v1.6.843 — Doctor explains NAS paths that were never mounted, and Tom AI fixes a box whose address changed - If SparkBox runs in a virtual machine or on a separate PC and you typed your NAS's own folder path into Media Root (like /volume1/your-share), downloads quietly stayed on the machine's own disk and never reached the NAS. sudo sparkbox doctor now says so in its Media Storage section: that a NAS path does not connect to the NAS by itself, how to check (findmnt -T <your folder), and that the share has to be mounted on the machine first (NFS or SMB, with nofail in /etc/fstab), then sudo sparkbox restart media. A real NAS is not affected. - When doctor finds that the box's saved network address is out of date (after a new address from your router or a move to another network), Tom AI can now offer the fix doctor itself prints, sudo sparkbox up, after you confirm. It refreshes the address and rebuilds the app links and the media wiring. sudo sparkbox doctor --fix offers the same. - When Tom AI runs a health check and still can't solve something, the case it hands to support now includes which checks failed, so support starts from what your box actually reported instead of from scratch. v1.6.844 — Tom AI tells you when a problem it couldn't fix passes after an update - When Tom AI couldn't fix a problem on your box and handed it to support, your box now re-checks that problem once after each update. If the check that found it now passes — and the app it is about is running — Tom AI tells you in the chat ("it passes now", or "fixed in SparkBox <version" when that release fixed it) and sends one notification. If the box can't be sure, it says nothing and checks again after the next update. The re-check runs on your box; the only thing that leaves it is a count of problems that passed after an update, sent with Tom AI's fix reports (setting SBAGENTTELEMETRY=0 turns those off). - Tom AI's health checks now record which checks passed, failed or were skipped, so its answers and the cases it hands to support are based on what your box actually reported. v1.6.845 — Tom AI links you straight to the VPN settings, and a safer support bundle - If your VPN is down, or was never set up, Tom AI now gives you the link that opens the VPN settings window from any dashboard layout, and walks you through filling it in: pick your VPN provider, then paste its WireGuard key and address (or import its WireGuard file), or use OpenVPN with its username and password, and save. Before, it only told you to check your VPN key, and the newer dashboard layouts have no VPN settings of their own. - sudo sparkbox diag saves its support bundle more safely: it can no longer write over a different file on your box. If it cannot save the bundle, it prints it on screen instead. v1.6.846 — Doctor shows why Nginx Proxy Manager is unhealthy - When Nginx Proxy Manager (your custom domains and HTTPS) shows as unhealthy and sudo sparkbox doctor finds nothing wrong with its ports or its folder, doctor now prints what its health check got back and Nginx Proxy Manager's last few lines, so you, Tom AI and support can see the reason. Your domain names and addresses are hidden in those lines. v1.6.847 — Media card clears after Manual Import, and Tom AI sees why an app won't start - When the Media card said a download "disappeared before it reached your library" and you filed it yourself with Manual Import in Sonarr or Radarr, the warning stayed up for up to two days. It now clears within a minute or two once the movie or episode is in your library. - When an app will not start at all (it shows as stopped with code 128), Tom AI now sees the reason Docker gave for refusing to start it, for example a drive or device the app needs is missing, instead of an empty log, so it can tell you what to fix. v1.6.848 — Jellyfin password reset turns a locked account back on - If Jellyfin said "You are not authorized to access the server at this time" for your SparkBox Jellyfin account, even with the right password, Jellyfin had switched that account off after too many wrong passwords in a row, and the Jellyfin password reset could not turn it back on. Now it does: update, then press Reset on the Jellyfin row in Settings → Passwords (or run sudo sparkbox reset-password jellyfin). You get a new password on that page and the account works again; your library is not touched. Afterwards, put the new password into any TV or phone app that signs in with that account. - sudo sparkbox doctor now says when Jellyfin has switched the account off, instead of saying the saved password is out of date. v1.6.849 — Jellyfin password reset works when it's your only admin account - If Jellyfin said "You are not authorized to access the server at this time" for your SparkBox Jellyfin account and the Jellyfin password reset then said there is no account called "sparkbox" or told you to use --resync, the reset could not get past it when the account was your only Jellyfin admin. Now it can: update, then run sudo sparkbox reset-password jellyfin in the box's terminal. Jellyfin stops for a few seconds while the account is switched back on (anyone watching presses play again), and you get a new password that works for Jellyfin and Seerr. Your library, your family's accounts and your settings are not touched, and a copy of Jellyfin's database is kept first. v1.6.850 — VPN setup says when the WireGuard Address is missing - If you set up Surfshark (or Mullvad, Windscribe, IVPN, FastestVPN, AirVPN or your own WireGuard VPN) by pasting only the private key, the VPN could not start and every download app stayed stopped. SparkBox now won't save the key without its WireGuard Address and tells you what to copy: the Address line from your provider's WireGuard config, which for Surfshark is 10.14.0.2/16. If your download apps are stopped for this reason right now, open Edit VPN, fill in WireGuard Address and save. Proton VPN and NordVPN don't need it. - When the VPN stops because that address is empty, the VPN alert and sudo sparkbox doctor now say so, instead of asking you to add /32 to an address you never typed. - When a fix Tom AI offers is declined or its confirmation times out, Tom's answer now keeps its explanation of what is wrong, with a note that the fix did not run, instead of only saying it skipped it. v1.6.851 — Tom AI starts your download apps again after you fix the VPN settings - If a restart of your download apps failed because the VPN could not start (for example IPVanish set up over WireGuard), and you then saved working VPN settings (IPVanish over OpenVPN with your username and password), Tom AI kept refusing to restart them for a week, saying it had already tried, and the apps stayed stopped. Tom AI now restarts them once your VPN settings have been saved since that failed try. If your download apps are stopped right now, ask Tom AI again after updating, or run sudo sparkbox restart media.