Complete application for automatic recording of Chaturbate and m3u8 streams

All-in-one Docker container to watch and automatically record HLS/m3u8 streams, with a modern web interface.

• Automatic recording of Chaturbate streams by username
• Modern web interface to control recordings
• Automatic detection when a user goes online
• Daily rotation of files (1 TS file per day)
• Direct m3u8 URL support for any type of stream
• Docker ready with docker-compose for Portainer/Umbrel
• Background auto-recording (no need to keep page open)
• Server-side storage (works in private browsing)
• Multilingual (French/English)
• GitOps updates (one-click update from GitHub)

• Recordings: /data/records/<person>/YYYY-MM-DD.ts
• HLS streaming: /data/sessions/<session_id>/
• Format: MPEG-TS compatible with all players (VLC, MPV, etc.)
• Models list: /data/models.json (server-side storage)

docker run -d \
  --name p-streamrec \
  -p 8080:8080 \
  -v ./data:/data \
  -e CB_RESOLVER_ENABLED=true \
  ghcr.io/raccommode/p-streamrec:latest

version: "3.8"
services:
  p-streamrec:
    image: ghcr.io/raccommode/p-streamrec:latest
    container_name: p-streamrec
    ports:
      - "8080:8080"
    volumes:
      - ./data:/data
    environment:
      - CB_RESOLVER_ENABLED=true
      - TZ=America/New_York
    restart: unless-stopped

If error "failed to load the compose file":

1. Verify that docker-compose.yml is present in the repo
2. Check that "Compose path" = docker-compose.yml
3. Check that branch = main

1. Add a model:

   • Click the + button
   • Enter a Chaturbate username (e.g., username)
   • Or paste a direct m3u8 URL
   • Click Add

2. Automatic recording:

   • Once a model is added, the system checks every 2 minutes
   • When the model goes online, recording starts automatically
   • No need to keep the page open!

3. Watch replays:

   • Click on a model card
   • Go to the Replays tab
   • Click on a recording to watch
   • Progress is automatically saved

4. Manage models:

   • Click on a model card to view details
   • Click the delete button to remove from list
   • Models are saved server-side (works in private browsing)

• The server checks every 2 minutes if your models are online
• Automatically starts recording when they go online
• Works 24/7 even if you close your browser
• Logs available in Docker/Portainer

• Automatically selects the highest quality available
• Manual quality selection if multiple streams available
• Options: Auto (best), 1080p, 720p, 480p, etc.

• Replay metadata cached for fast loading
• 30x faster on subsequent loads
• Automatic cache invalidation when files change

• One-click updates directly from the web interface
• Check for new versions from GitHub
• Apply updates without SSH or command line
• Automatic application restart after update
• Perfect for Portainer/Docker deployments

How to use:

1. Look for the GitOps button in the header (appears only if in a Git repository)
2. Click to see current version and available updates
3. If updates are available, click Apply Update
4. Application pulls latest changes and restarts automatically

Requirements:

• Application must be deployed as a Git clone
• Git must be available in the container
• Repository must have remote configured (origin)

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
uvicorn app.main:app --reload

Files are stored in /data/records/<username>/YYYY-MM-DD.ts

To play TS files:

• VLC: Open file directly
• MPV: mpv /path/to/file.ts
• FFmpeg convert to MP4:

  ffmpeg -i input.ts -c copy output.mp4

• Privacy respect: Use only for public content
• Storage: TS files can be large (~2-4 GB/hour)
• Bandwidth: Each stream uses approximately 1-3 Mbps
• CPU: Minimal usage (simple stream copy)

Current version: 2025.40.B

See version.json for changelog and release information.

• Use this software in compliance with laws and service ToS
• Does not bypass any technical protection measures
• Ensure you have the right to record the content
• Respect privacy and copyright laws