ChannelHoarder
Docker app from sworcery's Repository
ChannelHoarder is a self-hosted video channel archiver with a modern web UI.
Overview
Readme
View on GitHub
ChannelHoarder
Self-hosted video channel archiver with a modern web UI, designed for Plex-compatible output.
Runs as a single Docker container and automatically downloads new videos from your subscribed channels.
Features
- Multi-platform - Subscribe to channels (or YouTube playlists) from YouTube, Rumble, Twitch, Dailymotion, Vimeo, and Odysee by URL or @handle
- Automatic archiving - Scheduled scans detect new videos and queue them, with per-channel quality (up to 4K), preferred codec (H.264 or VP9), quality cutoff, title (keyword/regex) and duration filters, an exact start date, optional auto-delete of downloaded videos after N days, and per-channel storage paths
- Plex / Jellyfin / Emby ready - TV Show-style naming (seasons by year, chronological episodes) with customizable templates, NFO metadata, and poster/season art
- Sonarr-style episode management - Monitor/unmonitor, status icons, collapsible seasons, bulk actions, per-episode retry/re-download/rename/delete, and a channel-level Repair Dates action for videos saved with the wrong upload date
- Robust downloads - Queue that downloads one video at a time by default (Settings > General allows up to 5 at once), which you can order (newest, oldest, shortest, longest), with "download first" channels and a per-item "Move to top", rate limiting and live progress (speed/ETA) over WebSocket, standalone single-video downloads, import of existing files, subtitle and chapter support, and automatic Shorts / livestream filtering
- Smart authentication - Cookies (auto-sync via Windows exporter or Tampermonkey, or manual upload) with PO token fallback, optional YouTube Data API key, and a selectable yt-dlp player client
- Anti-detection - Configurable download delays, random jitter, and user-agent rotation
- Monitoring - Error classification with suggested fixes, diagnostic reports, searchable logs, and Telegram / Pushover / Discord notifications
- Self-contained - Single Docker container (web UI, download engine, PO token server, scheduler) with dark-mode UI, current yt-dlp on every image build, and config import/export
Contents
Quick Start · How It Works · File Organization · Configuration · Cookie Authentication · Supported Platforms · Reference · Troubleshooting
Quick Start
Docker Compose
services:
channelhoarder:
image: ghcr.io/sworcery/channelhoarder:latest
container_name: channelhoarder
ports:
- "8587:8000"
volumes:
- ./config:/config
- /path/to/your/media/youtube:/downloads
- ./cookies:/cookies
environment:
- PUID=1000
- PGID=1000
- TZ=America/New_York
restart: unless-stopped
Note: Set
PUID/PGIDto match your media server user so downloaded files have the correct ownership. On Unraid, this is typicallyPUID=99(nobody).
Then open http://your-server-ip:8587 in your browser. Update later with docker compose pull && docker compose up -d.
Requirements: Docker & Docker Compose; ~500 MB–1 GB disk per video at best quality (50 GB+ recommended); runs on Linux, macOS, Windows (Docker Desktop), Unraid, or TrueNAS.
First steps
- Go to Settings → Authentication and verify the PO token status reads Running
- (Optional) Add a YouTube Data API key for faster channel discovery
- (Optional) Configure Telegram, Pushover, or Discord notifications
- Go to Channels → Add Channel and paste a channel URL
- With Scan for videos after adding checked, the channel is scanned at once. With Auto-download videos checked, every video the channel has published is queued for download, unless you set a Download from date
Security note: ChannelHoarder has no built-in authentication - it is designed for trusted local networks. If exposing to the internet, place it behind a reverse proxy with authentication (e.g. Authelia, Nginx Proxy Manager).
How It Works
Adding channels
Show details
Paste a channel URL or @handle (e.g. https://www.youtube.com/@ChannelName) on the Channels page. ChannelHoarder fetches the channel metadata (name, description, thumbnail), adds it to your subscription list, and lets you set options in the Add dialog: Scan for videos after adding, Auto-download videos and Download from date. By default a channel downloads everything it has published unless a date is set. With Auto-download off, scans only list new videos and you queue the ones you want. Other per-channel options (quality, download folder and more) are under Edit on the channel page. To add many channels at once, use Bulk Import to paste a list or upload a CSV/text file, such as YouTube Takeout's subscriptions.csv.
Automatic scanning
Show details
- Each channel is scanned at its own random time, every 12 to 24 hours by default (set the interval under Settings > Scanning, and optionally limit scans to certain hours with a scan window on the same tab)
- Optionally, YouTube channels' RSS feeds can be checked every so many minutes (Settings > Scanning, off by default), and a channel is scanned as soon as its feed shows a new video. Checks stay under YouTube's daily feed limit, and interval scans keep running as the backup
- Each scan checks for new videos that haven't been downloaded yet and adds them to the download queue with
pendingstatus - If a YouTube Data API key is configured, it's used for faster and more reliable discovery; otherwise yt-dlp handles discovery directly
- With Auto-download off for a channel, scans only list new videos; they are not queued
- Manual scans (Scan Now per channel, Scan All Channels Now for all) run at any hour, even outside the scan window. A scan window must be at least 8 hours wide. Scheduled scans and feed checks stay inside it. The Dashboard says when scans are more than a day late
Downloads
Download pipeline
The Downloads page shows the queue, the history and the failed downloads. Pause stops new downloads while in-progress ones finish. Downloads run one at a time by default; Settings > General allows up to 5 at once. The pipeline works like this:
- Queue Processing - Every 30 seconds, the next queued video is picked up
- Rate Limiting - A configurable delay (default 10–30 seconds) with optional random jitter is applied between downloads
- Download - yt-dlp downloads the video with PO tokens, plus thumbnail and metadata. By default yt-dlp auto-selects a working player client; a specific client can be forced in Settings → Authentication
- Naming - Files are renamed to the Plex-compatible format with season/episode numbering
- Verification - Output files are verified to exist and the database is updated
- Progress - Real-time progress (speed, ETA, percentage) is broadcast via WebSocket to connected browsers
Importing existing videos
Show details
If you already have downloaded videos from a channel, you can import them instead of re-downloading:
- Open a channel's detail page and click Import Existing
- Enter the folder path (on the server) containing your video files
- ChannelHoarder scans the folder and fuzzy-matches filenames against undownloaded video entries
- Review the matches with confidence scores, select which to import
- Imported files are moved into the correct Plex-compatible directory structure
Important: Video filenames must contain the original video title for matching to work. Files named with only dates or generic names won't match.
Channel management
Show details
Open a channel to change how it downloads. The options apply to future scans and leave videos already queued alone:
- Quality - the target quality for new downloads, also the cutoff for upgrade searches, and a minimum quality. A video below the minimum isn't downloaded at once: the app waits and checks again after 1, 6 and 24 hours, then the video fails with a "Download anyway" button
- Codec - the preferred video codec. H.264 plays on the most devices but on YouTube is usually capped at 1080p; if the codec isn't available the best available format is used, and nothing is re-encoded
- Publish date - only download videos published on or after a date
- Title filter - include only matching titles, or exclude matching ones
- Delete after - remove downloaded files a set number of days after download. The videos stay listed as skipped and are not downloaded again, including after Force Rescan
- Auto-download - new videos found by scans are queued automatically. When off, they are only listed and you queue them yourself
- Scan this channel - when off, the channel is not scanned on schedule, by feed checks or by Scan All Channels Now. Scan Now still works
- Download first - queued videos from this channel download before other channels' videos
- Scan Now - scans the channel at once. A cooldown (Settings > Scanning) stops repeated clicks
- Force Rescan - rebuilds the records of monitored videos that are pending, failed or queued and have no file, then scans again. Downloaded, skipped, unmonitored and downloading videos are kept as they are. Nothing changes if the channel's video list can't be read
- Download Directory - Save & Move shows what will move before it moves. A plain Save makes the next scan move the files. Rescan the library in Plex or Jellyfin afterwards
Episode management
Show details
A channel's videos are listed as episodes. YouTube Shorts (on the Shorts tab, a /shorts/ URL, 60 seconds or less, or tagged #short) and livestreams or premieres have their own sections and are skipped unless both Settings > Media Management and the channel allow them. A minimum duration skips shorter videos; the maximum duration in Settings flags longer ones for review. Subtitles, chapters and SponsorBlock are set under Settings > Media Management.
Error handling and channel health
Show details
When a download fails, ChannelHoarder classifies the error and tells you the cause and what to do. The Failed tab on the Downloads page groups failures by cause and gives one tip per group; open a row's Details for the full explanation. The common causes:
- Rate limited, network error, download stalled, merge (ffmpeg) error, PO token failure. These retry on their own, up to Max Retries (Settings > General). Once the retries run out, fix the cause (raise the download delay, check the connection and free disk space, check the PO token server) and use Retry all.
- Disk full. The queue pauses until you free space on the downloads volume and press Resume.
- Quality too low. Only formats below the channel's minimum quality were offered. They retry after 1, 6 and 24 hours; use Download anyway to take the best quality now.
- Authentication required or age restricted. Upload fresh cookies in Settings > Authentication, then retry.
- Scan failed. The channel's video list could not be read. The scan tries again on its next run; see Settings > Diagnostics if it keeps failing.
- No usable format. Set the player client to Default (automatic) in Settings > Authentication, then retry.
- Dates could not be read (metadata degraded). Upload dates could not be fetched during a scan, usually from expired cookies. Those videos are left out, not saved with a wrong date. Upload fresh cookies and scan again; nothing is lost.
- Scheduled livestream. The video is a livestream or premiere that has not started. It is marked as a livestream and unmonitored; wait until it ends, then unmark it or turn on livestream downloads for the channel.
- Unavailable, private, removed or geo-blocked. The video itself is gone or blocked, so a retry will not help.
- yt-dlp outdated. Update the image; each build ships a current yt-dlp.
Each channel also shows a health indicator: Green (recent downloads succeeded), Yellow (some recent failures), or Red (most/all recent downloads failing, with the specific error reason shown).
File Organization
Directory layout & naming template
Videos are saved in Plex TV Show format. Each channel becomes a "show", each year becomes a "season", and videos are numbered as episodes in upload order:
/downloads/
Channel Name/
tvshow.nfo
poster.jpg
Season 2024/
S2024E001 - Video Title - 20240115 - [videoId].mp4
S2024E001 - Video Title - 20240115 - [videoId].nfo
S2024E002 - Another Video - 20240203 - [videoId].mp4
...
Season 2025/
S2025E001 - New Video - 20250112 - [videoId].mp4
...
The default naming template is:
{channel_name}/Season {season}/S{season}E{episode} - {title} - {upload_date} - [{video_id}]
Available template variables: {channel_name}, {season}, {episode}, {title}, {upload_date}, {video_id}
Plex library setup
Create a separate Plex library for your YouTube channels - do not mix them with your regular TV shows library.
- In Plex, click + to add a new library
- Select TV Shows as the type
- Point it at your ChannelHoarder downloads folder
- Under Advanced, set the agent to Personal Media Shows
- Check Use local assets
- Save and scan
Why a separate library? The Personal Media Shows agent reads episode titles, descriptions, and artwork from the NFO files that ChannelHoarder generates. The default Plex TV agent ignores NFO files and only looks up online databases (TVDB/TMDB), which won't find YouTube channels. Using Personal Media Shows in your regular TV library would break metadata for shows from Sonarr/etc.
If you already added channels to an existing TV library, move them to the new library and unmatch each show for the NFO data to take effect.
Configuration
Environment variables
| Variable | Default | Description |
|---|---|---|
PUID |
1000 |
User ID for file permissions |
PGID |
1000 |
Group ID for file permissions |
TZ |
America/New_York |
Timezone for scheduling |
LOG_LEVEL |
info |
Logging level (debug, info, warning, error). Starting value only; change it in Settings > General |
YOUTUBE_API_KEY |
(empty) | Optional YouTube Data API v3 key for reliable channel discovery |
POT_SERVER_ENABLED |
true |
Enable the built-in PO token server |
POT_SERVER_URL |
http://127.0.0.1:4416 |
PO token server address (internal to container) |
CONFIG_DIR |
/config |
Configuration and database storage path |
DOWNLOAD_DIR |
/downloads |
Video output directory |
EXTRA_DOWNLOAD_DIRS |
(empty) | Comma-separated additional allowed download paths (e.g. /media,/cartoons) |
TEMP_DIR |
(empty) | Optional folder for in-progress downloads, such as a volume on an SSD. Finished videos are then moved into their download folder, so a slow or external download drive only gets the finished file. Empty means downloads happen inside the download folder itself |
COOKIE_WATCH_DIR |
/cookies |
Directory watched for cookie file updates |
DEFAULT_QUALITY |
best |
Default quality for new channels. Starting value only; change it in Settings > General |
MAX_CONCURRENT_DOWNLOADS |
1 |
Maximum simultaneous downloads. Starting value only; change it in Settings > General |
MAX_RETRIES |
3 |
Retry attempts for failed downloads. Starting value only; change it in Settings > General |
DOWNLOAD_DELAY_MIN |
10 |
Minimum delay between downloads (seconds). Starting value only; change it in Settings > Anti-Detection |
DOWNLOAD_DELAY_MAX |
30 |
Maximum delay between downloads (seconds). Starting value only; change it in Settings > Anti-Detection |
JITTER_ENABLED |
true |
Add random 0–10s jitter between downloads. Starting value only; change it in Settings > Anti-Detection |
USER_AGENT_ROTATION |
true |
Rotate browser user-agent strings. Starting value only; change it in Settings > Anti-Detection |
On the first start of 2.0, any MAX_RETRIES, USER_AGENT_ROTATION and LOG_LEVEL values you set are copied into Settings once, so nothing changes. After that, change them in Settings.
YouTube Data API key (optional)
How to get a key
A YouTube Data API key improves channel scanning reliability and speed. Without it, yt-dlp handles discovery directly, which works but can be slower.
To get a free API key:
- Go to console.cloud.google.com and sign in with your Google account
- Click Select a project at the top, then New Project. Name it anything (e.g. "ChannelHoarder") and click Create
- Make sure your new project is selected, then go to APIs & Services > Library
- Search for YouTube Data API v3 and click on it
- Click Enable
- Go to APIs & Services > Credentials
- Click + Create Credentials > API key
- Copy the API key that appears
- In ChannelHoarder, go to Settings > Authentication and paste the key, or set the
YOUTUBE_API_KEYenvironment variable in your container config
The free tier provides 10,000 quota units per day, which is enough for most personal use. No billing account is required.
Settings in the web UI
Show details
All settings are on the Settings page, in seven tabs. Type in the search box at the top to find any setting by name; choosing a result opens its tab and highlights it. Changes on the General, Scanning, Media Management, Anti-Detection and Notifications tabs wait for the Save button in the bar that appears at the bottom; Discard puts the saved values back. Switching tabs or leaving the page with unsaved changes asks first. On the Authentication and Diagnostics tabs, and for Export/Import Config, Scan All Channels Now, naming Preview and the notification test buttons, the action happens when you press its button.
- General - Default quality, max concurrent downloads, max retries, log level, system info, the bundled yt-dlp version, config export/import
- Scanning - How often each channel is scanned, RSS feed checks for new YouTube uploads, the scan window (limit scans to certain hours), the Scan Now cooldown, and a Scan All Channels Now button
- Media Management - Output filename template with live preview (
{channel_name},{season},{episode},{title},{upload_date},{video_id}), Shorts and livestream opt-in, subtitles, chapters, SponsorBlock, file permissions, max video duration filter - Authentication - PO token status, cookie upload/status, browser cookie sync (Tampermonkey), YouTube API key, player client strategy
- Anti-Detection - Download delay range, random jitter, user-agent rotation, scan jitter
- Notifications - Telegram, Pushover, and Discord configuration with test buttons, per-event toggles
- Diagnostics - System overview and the diagnostic report, error logs, system logs, the Library Check and Shared Folders repair tools
Only variables marked "Starting value only" in the table above are starting values; once a value is saved in Settings, its variable no longer applies. PUID, PGID, TZ, the directories, the POT_SERVER_* variables and COOKIE_WATCH_DIR apply as set. YOUTUBE_API_KEY is different: if it is set, it is used when the container starts, even over a key saved in Settings.
Anti-detection
Show details
Random delays between downloads (default 10 to 30 seconds), optional random jitter (0 to 10 extra seconds), rotating user-agent strings and a random pause between channel scans keep traffic from looking like a script. All of it is under Settings > Anti-Detection.
Notifications
Show details
Under Settings > Notifications, add a Telegram bot (token and chat ID), Pushover (application token and user key) and/or a Discord webhook URL. Save first, then use the test buttons, which send through the saved settings. Per-event toggles choose what triggers a message: Download Complete, Download Failed, Cookies Expired, Cookies Refreshed, Health Alerts, Critical Alerts, Long Video Review Required, File Move Complete, Channel Scan Complete, Channel Scan Failed and Quality Upgrade.
Cookie Authentication
PO tokens
Show details
YouTube asks clients to prove they are a real browser. The built-in PO token server makes those proofs automatically, so downloads work without any setup. Settings > Authentication shows its status: Running, Not responding or Turned off. If it is not responding, restart the container; uploading cookies works as a fallback. When you also upload cookies, the cookies take priority.
Player client
Show details
The player client is the YouTube API yt-dlp uses to list a video's formats. Leave it on Default (automatic): yt-dlp picks a working client and adapts when YouTube changes. Override it only if one specific video fails with a format error; ios and web_safari are good fallbacks. The choice applies as soon as you select it.
Cookies (optional)
Three methods
Cookies are optional but can help if you have YouTube Premium, need to access region-specific content, or want to download premium/authenticated Rumble content. The same cookies.txt covers all platforms - cookies are matched per-domain, so you can include YouTube and Rumble cookies in one file. There are three ways to provide cookies:
Option 1: Automatic Cookie Sync (Recommended)
The Windows cookie exporter reads cookies directly from Firefox's database, including HttpOnly cookies that browser scripts cannot access.
- Install Python 3.10+ and pycryptodome on a Windows machine or VM
- Install Firefox and log into YouTube
- Copy the
tools/folder from this repository to the machine - Edit
cookie_exporter.ini:- Set
server_urlto your ChannelHoarder address (e.g.,http://your-server:8587) - Optionally set
profileif using a non-default Firefox profile - Optionally set
domainsto customize which cookie domains to export (default:.youtube.com, .google.com, .rumble.com- log into Rumble in Firefox too if you want premium Rumble content) - Optionally set
refresh_urlto change the page loaded before export (default:https://www.youtube.com) - Optionally set
refresh_waitto change seconds to wait for page load (default:20)
- Set
- Run
python cookie_exporter.py --dry-runto test - Run
install_task.batas administrator to set up automatic export every 30 minutes - Every run: opens Firefox to YouTube (refreshes cookies) → closes Firefox → reads the cookie DB → pushes to ChannelHoarder via API
Option 2: Browser Cookie Sync (Tampermonkey)
- Install the Tampermonkey browser extension
- Go to Settings > Authentication in ChannelHoarder and click Install Tampermonkey Script
- The script comes pre-configured with your server address and exports cookies each time you load a YouTube page (with a 5-minute cooldown between exports)
- Note: Cannot access HttpOnly cookies (SID, HSID, SSID) - Option 1 is more complete
Option 3: Manual Upload
- Use a cookie export extension (e.g., "Get cookies.txt LOCALLY")
- Navigate to youtube.com while logged in and export cookies
- Upload the file via Settings > Authentication in the web UI
Cookies expire periodically. Options 1 and 2 handle re-export automatically.
Supported Platforms
| Platform | Channel Discovery | Video Download | API Support |
|---|---|---|---|
| YouTube | RSS + API + yt-dlp | yt-dlp | YouTube Data API v3 |
| Rumble | yt-dlp | yt-dlp | No |
| Twitch | yt-dlp | yt-dlp | No |
| Dailymotion | yt-dlp | yt-dlp | No |
| Vimeo | yt-dlp | yt-dlp | No |
| Odysee | yt-dlp | yt-dlp | No |
Reference
Tech stack
Tech stack & volumes
- Backend: Python 3.12, FastAPI, SQLAlchemy (async), APScheduler, yt-dlp
- Frontend: React 18, TypeScript, Tailwind CSS, TanStack Query, React Router
- Database: SQLite (WAL mode)
- Auth: bgutil-ytdlp-pot-provider (PO tokens), optional cookies.txt
- Docker: Multi-stage build with Python, Node.js, ffmpeg, Deno, and the PO token provider
Volumes:
| Path | Purpose |
|---|---|
/config |
Database, settings, API key, optional cookies.txt |
/downloads |
Downloaded videos, thumbnails, and metadata (point Plex here) |
/cookies |
Optional: watched directory for cookie file updates |
API
Show details
ChannelHoarder exposes a REST API under /api/v1/ plus a WebSocket for live progress. See docs/API.md for the full endpoint list, or browse the interactive Swagger UI at /docs on a running instance.
Dashboard widgets: GET /api/v1/dashboard/stats returns channel, download, queue and storage counts as JSON, so dashboard apps can show them. For Homepage, use its Custom API widget:
- ChannelHoarder:
href: http://your-server:8587
widget:
type: customapi
url: http://your-server:8587/api/v1/dashboard/stats
mappings:
- field: total_channels
label: Channels
- field: total_downloaded
label: Downloaded
- field: queue_length
label: Queue
- field: storage_used_bytes
label: Storage
format: bytes
Known limitations
Show details
- YouTube Shorts - Shorts exist only on YouTube. A YouTube video counts as a Short if it is on the channel's Shorts tab, has a
/shorts/URL, runs 60 seconds or less, or has the #short or #shorts tag. Shorts are skipped by default (toggleable per channel). Short clips on other sites download like any video. To get a skipped one back, open its menu on the channel page, choose Unmark as Short, then queue it. The choice sticks - Livestreams / long videos - Videos over the max duration setting are skipped by default
- Geo-blocked content - Cannot download region-restricted videos
- Private / deleted videos - Automatically skipped if no longer available
- No built-in auth - Designed for trusted local networks; use a reverse proxy if internet-facing
- Rate limiting - YouTube may throttle requests; the PO token server and configurable delays help mitigate this
Troubleshooting
Library repair
Library repair (Settings > Diagnostics)
Two tools fix problems older versions could leave behind. Both show what they found before changing anything, and neither overwrites a file.
Library Check. Before 1.9.50, a scan could move a downloaded file onto a newer video with a similar title (for example "Week 11" and "Week 12") and rename it. Run the check once if you have channels with numbered titles. It lists:
- files that are certainly filed under the wrong video, from the file's embedded source address or the
.info.jsonbeside it. "Fix all certain" moves each to the video it really belongs to and queues the wrongly filed video to download again; - likely mismatches, found from the file's length. Fix these one at a time;
- downloaded videos whose file is gone. Re-download or skip them;
- files to check by hand.
Shared Folders. Before 1.9.50, two channels with the same name (for example one creator on YouTube and on Rumble) shared one folder. New channels get their own folder, such as "Creator (Rumble)". For existing ones, the dashboard shows a notice, and this tool moves the channel you pick into a folder of its own. Files the other channel uses, shared artwork and files neither channel knows about stay where they are.
Common issues and fixes
Show details
Queue not downloading / stuck
- Look for a Queue paused banner on the Downloads page. It gives the reason: paused by you, disk full, or cookies expired. Press Resume (disk full needs free space first; expired cookies need a fresh upload)
- The strip across the top of the app lists anything that needs you: a full or nearly full disk, expired cookies, a PO token server that is not responding, a paused queue, channels that need attention (channels that keep failing), channels overdue for a scan, and the shared-folders notice (its button, Separate them, opens the fix). Each has a button that goes to the fix
- If a scan window is set (Settings → Scanning), scheduled scans and feed checks only run inside it, so channels can wait for the next window. Scan Now and Scan All Channels Now run at any hour. The Dashboard says when scans are more than a day late
- A row reading Waiting for higher quality is held back on purpose: only low formats were offered, and it retries after 1, 6 and 24 hours. Download anyway skips the wait
- A row reading Waiting: channel is finishing its current download first (or is being moved, is being scanned, is busy with another operation) starts by itself once that work ends
- Check Settings → Authentication: the PO token status must read Running. If it says Not responding, restart the container; uploading cookies works as a fallback
- Check Settings → Diagnostics → System Logs for error details
- If cookies expired, upload fresh ones via Settings → Authentication
"Sign in to confirm you're not a bot"
- Your session has been flagged - try uploading fresh cookies
- Adding a YouTube Data API key improves reliability
- Wait 1–2 hours if recently rate-limited
Downloads failing with "Requested format is not available"
- YouTube restricts certain player clients (DRM-protected or URL-less SABR formats), leaving nothing downloadable
- Set the player client to Default (automatic) in Settings → Authentication so yt-dlp picks a working client and adapts to YouTube's changes
- If a specific video still fails, try
iosorweb_safarifrom the same dropdown - Make sure the image is up to date (each build ships current yt-dlp)
Plex not showing downloaded videos
- Ensure your
/downloadspath is added as a TV Shows library in Plex - Trigger a library scan: Plex → Libraries → Scan Library Files
- Check that
PUID/PGIDmatch your Plex user so files are readable
Port conflict on 8587
- Change the host-side port:
"8590:8000"- then access via:8590
Container writes files as root
- Ensure
PUIDandPGIDare set correctly in your compose/template - The internal port is always
8000; the host port is configurable
Support
For bugs or feature requests, open a GitHub issue and include the diagnostic report from Settings → Diagnostics → System Overview → Copy Diagnostic Report.
License
MIT License - free to use, modify, and distribute with attribution. See LICENSE for details.
Categories
Related apps
Explore more like this
Explore allLinks
Details
ghcr.io/sworcery/channelhoarder:latestRuntime arguments
- Web UI
http://[IP]:[PORT:8000]- Network
bridge- Privileged
- false
Template configuration
Web interface port. Access the UI at http://your-server-ip:8587
- Target
- 8000
- Default
- 8587
- Value
- 8587
Configuration, database, and settings storage
- Target
- /config
- Default
- /mnt/user/appdata/channelhoarder
- Value
- /mnt/user/appdata/channelhoarder
Video download directory. Point your Plex TV Shows library here.
- Target
- /downloads
- Default
- /mnt/user/data/media/youtube
- Value
- /mnt/user/data/media/youtube
Optional: drop an updated cookies.txt here and it will be picked up automatically
- Target
- /cookies
- Default
- /mnt/user/appdata/channelhoarder/cookies
Optional: additional host path for per-channel custom download directories outside /downloads. To download here you must also list the container path (/media) in the 'Extra Download Dirs' variable below.
- Target
- /media
Comma-separated container paths (besides /downloads) that channels are allowed to download into. Each must match the container-side target of an extra path you mapped above - e.g. /media for the Extra Media Path, or add your own like /media,/playlists.
- Target
- EXTRA_DOWNLOAD_DIRS
- Default
- /media
- Value
- /media
Optional: a host folder on a fast drive (for example an SSD or the cache pool) where videos are downloaded before being moved into their download folder. Useful when the download folder is on a slow or external drive. To use it you must also set the 'Temp Dir' variable below to /temp.
- Target
- /temp
Container path for in-progress downloads. Leave empty to download inside the download folder itself (the default). Set to /temp to use the Temp Download Path above.
- Target
- TEMP_DIR
User ID for file permissions. Use 99 for nobody on Unraid.
- Default
- 99
- Value
- 99
Group ID for file permissions. Use 100 for users on Unraid.
- Default
- 100
- Value
- 100
Timezone (e.g. America/New_York, America/Chicago, Europe/London)
- Default
- America/New_York
- Value
- America/New_York
Logging level (debug, info, warning, error). Starting value only; change it in Settings > General
- Default
- info
- Value
- info
Optional: YouTube Data API v3 key for faster and more reliable channel discovery
- Target
- YOUTUBE_API_KEY
Enable built-in PO token server for YouTube authentication (recommended)
- Target
- POT_SERVER_ENABLED
- Default
- true
- Value
- true