Pachelarr

A smart caching proxy between Prowlarr and debrid services that dramatically improves torrent search results by prioritizing cached content and enriching metadata.

What It Does

Pachelarr sits between your media management tools (Radarr/Sonarr) and Prowlarr, intercepting torrent searches to:

Check debrid cache status - Queries your debrid service (currently Torbox) to identify which torrents are already cached
Boost cached results - Adds a configurable seeder boost (default: 10,000) to cached torrents, making them appear at the top of search results
Enrich metadata - Optionally scrapes trackers for real-time seeders/leechers data
Resolve IDs to titles - Uses TMDB API to convert IMDb/TVDB/TMDB IDs into searchable text queries, fixing compatibility with public indexers that don't support ID-based searches

The Problem It Solves

1. Debrid Service Integration

Problem: Radarr/Sonarr can't check if torrents are cached before downloading, often grabbing uncached torrents that take hours to download to your debrid service.

Solution: Pachelarr checks cache status in real-time and boosts cached torrents to the top of results. Your media manager will almost always pick instantly-available cached content.

2. ID-Only Search Failures

Problem: When Radarr/Sonarr search by IMDb/TVDB ID alone (without text), public torrent indexers return wrong results because they don't support native ID-based searches.

Example: Searching for "Elf (2003)" with IMDb ID tt0319343 returns random movies like "Emperor's New Groove" and "Fraggle Rock."

Solution: Pachelarr automatically looks up the movie/TV title from the ID using TMDB API and adds it to the search, ensuring accurate results.

3. Poor Metadata Quality

Problem: Many public indexers don't provide accurate seeder/leecher counts, making it hard to judge torrent health.

Solution: Enable optional UDP tracker scraping to get real-time stats directly from trackers.

Compatibility

Media Management (Clients)

✅ Radarr (all versions)
✅ Sonarr (all versions)
✅ Any Torznab-compatible client

Indexer Aggregators

✅ Prowlarr (required)

Debrid Services

✅ Torbox (fully supported)
🔄 Real-Debrid, AllDebrid, Premiumize (coming soon)

Note: Pachelarr currently only works with Torbox but will be updated to support other popular debrid providers in future releases.

How It Works

┌─────────┐     ┌───────────┐     ┌──────────┐     ┌─────────┐
│ Radarr/ │────▶│ Pachelarr │────▶│ Prowlarr │────▶│ Public  │
│ Sonarr  │◀────│           │◀────│          │◀────│ Indexers│
└─────────┘     └─────┬─────┘     └──────────┘     └─────────┘
                      │
                      │ Check cache status
                      ▼
                ┌──────────┐
                │  Torbox  │
                │   API    │
                └──────────┘
                      │
                      │ Lookup titles from IDs
                      ▼
                ┌──────────┐
                │   TMDB   │
                │   API    │
                └──────────┘

Request Flow:

Radarr/Sonarr sends search request to Pachelarr (configured as a Torznab indexer)
If search contains only IDs, Pachelarr looks up the title via TMDB
Pachelarr forwards enriched search to Prowlarr
Prowlarr queries all enabled public indexers
Pachelarr receives results and extracts info hashes
Pachelarr checks which torrents are cached in Torbox
Cached torrents get massive seeder boost (10,000+ seeders added)
Results returned to Radarr/Sonarr with cached content at the top
Radarr/Sonarr picks the "best" torrent (almost always cached)

Installation

Prerequisites

Docker and Docker Compose
Prowlarr instance with configured indexers
Torbox account and API key
TMDB API key (free, required for ID-based searches)

Using Docker Image Directly

You can also pull and run the pre-built Docker image without cloning the repository:

# Pull the latest image
docker pull ghcr.io/northernpowerhouse/pachelarr:latest

# Run with docker run
docker run -d \
  --name pachelarr \
  -p 6800:6800 \
  -e PROWLARR_URL=http://your-prowlarr-host:9696 \
  -e PROWLARR_API_KEY=your_prowlarr_api_key_here \
  -e TORBOX_API_KEY=your_torbox_api_key_here \
  -e PACHELARR_API_KEY=your_pachelarr_api_key_here \
  -e TMDB_API_KEY=your_tmdb_api_key_here \
  --restart unless-stopped \
  ghcr.io/northernpowerhouse/pachelarr:latest

Using Docker Compose (Recommended)

Clone or download this repository
Download the docker-compose.yml

You can download the compose file directly or copy it from the repository:

# Download directly
curl -O https://raw.githubusercontent.com/northernpowerhouse/pachelarr/main/docker-compose.yml

# Or if you prefer wget
wget https://raw.githubusercontent.com/northernpowerhouse/pachelarr/main/docker-compose.yml

Or copy this into your docker-compose.yml:

Click to expand docker-compose.yml

services:
  pachelarr:
    image: ghcr.io/northernpowerhouse/pachelarr:latest
    container_name: pachelarr
    ports:
      - "6800:6800"
    environment:
      # === REQUIRED CONFIGURATION ===
      # Prowlarr connection settings
      - PROWLARR_URL=http://your-prowlarr-host:9696
      - PROWLARR_API_KEY=your_prowlarr_api_key_here
      
      # Torbox API key for torrent caching
      - TORBOX_API_KEY=your_torbox_api_key_here
      
      # Pachelarr API key (used by Radarr/Sonarr to authenticate)
      - PACHELARR_API_KEY=your_pachelarr_api_key_here
      
      # === PACHELARR SETTINGS ===
      # Port to listen on (default: 8080)
      - PACHELARR_PORT=6800
      
      # Log level: DEBUG, INFO, WARNING, ERROR (default: INFO)
      - PACHELARR_LOG_LEVEL=INFO
      
      # Seeders boost value added to cached torrents (default: 10000)
      - PACHELARR_SEEDERS_BOOST=10000
      
      # Fallback query for category-only requests (improves Sonarr "Test" behavior)
      # Set to empty string to disable (default: "")
      - PACHELARR_TEST_FALLBACK_QUERY=
      
      # === TMDB CONFIGURATION ===
      # TMDB API key for looking up movie/TV titles from IMDb/TVDB/TMDB IDs
      # Get a free key at: https://www.themoviedb.org/settings/api
      # Without this, ID-only searches (common with Radarr/Sonarr) will fail
      - TMDB_API_KEY=your_tmdb_api_key_here
      
      # === TORBOX SETTINGS ===
      # Torbox API endpoint for checking cached torrents
      - TORBOX_CHECK_URL=https://api.torbox.app/v1/api/torrents/checkcached
      
      # Number of hashes to check per Torbox API request (default: 100, max: 100)
      - TORBOX_CHUNK_SIZE=100
      
      # Maximum retry attempts for Torbox API failures (default: 3)
      - TORBOX_MAX_RETRIES=3
      
      # Backoff delay in seconds between retries (default: 0.5)
      - TORBOX_RETRY_BACKOFF=0.5
      
      # === TRACKER SCRAPING SETTINGS ===
      # Enable direct UDP tracker scraping for seeders/leechers (default: false)
      # Warning: Enables direct contact with public trackers
      - TRACKER_SCRAPE_ENABLED=false
      
      # Number of concurrent tracker scrape requests (default: 4)
      - TRACKER_SCRAPE_CONCURRENCY=4
      
      # Timeout in seconds for tracker scrape requests (default: 5.0)
      - TRACKER_SCRAPE_TIMEOUT=5.0
      
      # Number of info hashes to scrape per tracker request (default: 50)
      - TRACKER_SCRAPE_BATCH_SIZE=50
    restart: unless-stopped

For Docker Compose GUI users (Portainer, Dockge, etc.):

Copy the compose content above
Paste it into your Docker Compose editor
Update the environment variables with your API keys
Deploy the stack

git clone https://github.com/northernpowerhouse/pachelarr.git
cd pachelarr

Clone or download this repository
Download the docker-compose.yml

You can download the compose file directly or copy it from the repository:

# Download directly
curl -O https://raw.githubusercontent.com/northernpowerhouse/pachelarr/main/docker-compose.yml

# Or if you prefer wget
wget https://raw.githubusercontent.com/northernpowerhouse/pachelarr/main/docker-compose.yml

Or copy this into your docker-compose.yml:

Click to expand docker-compose.yml

services:
  pachelarr:
    image: ghcr.io/northernpowerhouse/pachelarr:latest
    container_name: pachelarr
    ports:
      - "6800:6800"
    environment:
      # === REQUIRED CONFIGURATION ===
      # Prowlarr connection settings
      - PROWLARR_URL=http://your-prowlarr-host:9696
      - PROWLARR_API_KEY=your_prowlarr_api_key_here
      
      # Torbox API key for torrent caching
      - TORBOX_API_KEY=your_torbox_api_key_here
      
      # Pachelarr API key (used by Radarr/Sonarr to authenticate)
      - PACHELARR_API_KEY=your_pachelarr_api_key_here
      
      # === PACHELARR SETTINGS ===
      # Port to listen on (default: 8080)
      - PACHELARR_PORT=6800
      
      # Log level: DEBUG, INFO, WARNING, ERROR (default: INFO)
      - PACHELARR_LOG_LEVEL=INFO
      
      # Seeders boost value added to cached torrents (default: 10000)
      - PACHELARR_SEEDERS_BOOST=10000
      
      # Fallback query for category-only requests (improves Sonarr "Test" behavior)
      # Set to empty string to disable (default: "")
      - PACHELARR_TEST_FALLBACK_QUERY=
      
      # === TMDB CONFIGURATION ===
      # TMDB API key for looking up movie/TV titles from IMDb/TVDB/TMDB IDs
      # Get a free key at: https://www.themoviedb.org/settings/api
      # Without this, ID-only searches (common with Radarr/Sonarr) will fail
      - TMDB_API_KEY=your_tmdb_api_key_here
      
      # === TORBOX SETTINGS ===
      # Torbox API endpoint for checking cached torrents
      - TORBOX_CHECK_URL=https://api.torbox.app/v1/api/torrents/checkcached
      
      # Number of hashes to check per Torbox API request (default: 100, max: 100)
      - TORBOX_CHUNK_SIZE=100
      
      # Maximum retry attempts for Torbox API failures (default: 3)
      - TORBOX_MAX_RETRIES=3
      
      # Backoff delay in seconds between retries (default: 0.5)
      - TORBOX_RETRY_BACKOFF=0.5
      
      # === TRACKER SCRAPING SETTINGS ===
      # Enable direct UDP tracker scraping for seeders/leechers (default: false)
      # Warning: Enables direct contact with public trackers
      - TRACKER_SCRAPE_ENABLED=false
      
      # Number of concurrent tracker scrape requests (default: 4)
      - TRACKER_SCRAPE_CONCURRENCY=4
      
      # Timeout in seconds for tracker scrape requests (default: 5.0)
      - TRACKER_SCRAPE_TIMEOUT=5.0
      
      # Number of info hashes to scrape per tracker request (default: 50)
      - TRACKER_SCRAPE_BATCH_SIZE=50
    restart: unless-stopped

For Docker Compose GUI users (Portainer, Dockge, etc.):

Copy the compose content above
Paste it into your Docker Compose editor
Update the environment variables with your API keys
Deploy the stack

git clone https://github.com/northernpowerhouse/pachelarr.git
cd pachelarr

Get your API keys
- Torbox API Key: https://torbox.app/settings (under API section)
- Prowlarr API Key: Prowlarr → Settings → General → API Key
- TMDB API Key: https://www.themoviedb.org/settings/api (free account required)
Configure environment variables

Edit docker-compose.yml and update these required values:

- PROWLARR_URL=http://your-prowlarr-host:9696
- PROWLARR_API_KEY=your_prowlarr_api_key_here
- TORBOX_API_KEY=your_torbox_api_key_here
- TMDB_API_KEY=your_tmdb_api_key_here
- PACHELARR_API_KEY=your_pachelarr_api_key_here  # Any random string

Start the service

docker compose up -d

Configure Radarr/Sonarr

Add Pachelarr as a Torznab indexer:

URL: http://pachelarr-host:6800/api
API Key: Whatever you set for PACHELARR_API_KEY
Categories:
- Movies: 2000,2010,2020,2030,2040,2045,2050,2060,2070,2080
- TV: 5000,5010,5020,5030,5040,5045,5050,5060,5070,5080

Configuration

All configuration is done via environment variables in docker-compose.yml.

Required Settings

Variable	Description	Example
`PROWLARR_URL`	Prowlarr instance URL	`http://192.168.1.100:9696`
`PROWLARR_API_KEY`	Prowlarr API key	`abc123...`
`TORBOX_API_KEY`	Torbox API key	`xyz789...`
`PACHELARR_API_KEY`	API key for Radarr/Sonarr to authenticate	Any string
`TMDB_API_KEY`	TMDB API key for ID lookups	Get at themoviedb.org

Optional Settings

Pachelarr Settings

Variable	Default	Description
`PACHELARR_PORT`	`8080`	Port to listen on
`PACHELARR_LOG_LEVEL`	`INFO`	Log verbosity: DEBUG, INFO, WARNING, ERROR
`PACHELARR_SEEDERS_BOOST`	`10000`	Seeders added to cached torrents
`PACHELARR_TEST_FALLBACK_QUERY`	`""`	Fallback query for category-only searches (improves Sonarr "Test" button)

Torbox Settings

Variable	Default	Description
`TORBOX_CHECK_URL`	`https://api.torbox.app/v1/api/torrents/checkcached`	Torbox cache check endpoint
`TORBOX_CHUNK_SIZE`	`100`	Hashes per API request (max: 100)
`TORBOX_MAX_RETRIES`	`3`	Retry attempts on failure
`TORBOX_RETRY_BACKOFF`	`0.5`	Seconds between retries

Tracker Scraping Settings

Variable	Default	Description
`TRACKER_SCRAPE_ENABLED`	`false`	Enable UDP tracker scraping for real seeders/leechers
`TRACKER_SCRAPE_CONCURRENCY`	`4`	Concurrent scrape requests
`TRACKER_SCRAPE_TIMEOUT`	`5.0`	Timeout per scrape request (seconds)
`TRACKER_SCRAPE_BATCH_SIZE`	`50`	Info hashes per scrape request

⚠️ Warning: Enabling tracker scraping makes direct UDP connections to public trackers. Use at your own discretion.

Features

🚀 Cache-First Results

Automatically prioritizes torrents that are already cached on your debrid service, giving you instant downloads instead of waiting hours for uncached content.

🎯 Smart ID Resolution

Converts IMDb/TVDB/TMDB IDs to searchable titles using TMDB API, fixing broken ID-only searches that plague public indexers.

Supported ID types:

IMDb IDs (movies & TV)
TVDB IDs (TV shows)
TMDB IDs (movies & TV)
TVRage IDs (TV shows, legacy)

📊 Real-Time Metadata

Optional UDP tracker scraping provides accurate, real-time seeder and leecher counts directly from BitTorrent trackers.

🔄 Transparent Proxy

Works seamlessly with Radarr/Sonarr - no modifications needed to existing workflows. Just add as a Torznab indexer.

⚡ Performance Optimized

Bulk hash checking (up to 100 hashes per API call)
Concurrent tracker scraping
Configurable timeouts and retry logic
Request caching for faster repeated searches

Usage Examples

Typical Radarr Search

User adds "Elf (2003)" to Radarr
Radarr searches Pachelarr with imdbid=0319343
Pachelarr looks up "Elf (2003)" from TMDB
Pachelarr searches Prowlarr with "Elf 2003" + IMDb ID
Prowlarr returns 200+ torrents from public indexers
Pachelarr checks which are cached in Torbox (e.g., 25 cached)
Cached torrents get +10,000 seeders
Radarr sees cached 4K REMUX with 10,500 seeders at top
Radarr grabs cached torrent → instant download from Torbox

Manual Testing

Test the cache checking:

curl "http://localhost:6800/api?t=movie&cat=2000&imdbid=0319343&apikey=your_api_key"

Look for [CACHED] prefix in results - these torrents are instantly available.

Troubleshooting

No cached results appearing

Verify TORBOX_API_KEY is correct
Check logs: docker logs pachelarr-pachelarr-1
Ensure torrents exist in Torbox cache (check Torbox directly)

Wrong search results

Ensure TMDB_API_KEY is set and valid
Check logs for "Successfully looked up movie via TMDB"
Verify Prowlarr has working indexers

Radarr/Sonarr can't connect

Confirm PACHELARR_API_KEY matches in both places
Check port 6800 is accessible
Verify container is running: docker ps

Enable debug logging

- PACHELARR_LOG_LEVEL=DEBUG

Then restart: docker compose restart

Performance Considerations

TMDB API Rate Limits

Free tier: 40 requests per 10 seconds
Pachelarr usage: 1 request per ID-only search
More than sufficient for typical Radarr/Sonarr usage

Torbox API Limits

Check up to 100 hashes per request
Typical search checks 50-200 torrents (1-2 API calls)
No known rate limits for personal use

Tracker Scraping

Adds 1-3 seconds latency per search when enabled
Recommended for users who need accurate seeder counts
Disable if speed is more important than metadata accuracy

Privacy & Security

All API keys stored in environment variables (not in code)
No telemetry or analytics
TMDB lookups expose searched titles to TMDB servers
Tracker scraping exposes your IP to public BitTorrent trackers
Use with VPN if privacy is a concern

Contributing

Contributions welcome! Areas for improvement:

Add support for Real-Debrid, AllDebrid, Premiumize
Implement proper caching layer (Redis/SQLite)
Add web UI for configuration and stats
Support for more ID types (Trakt, TVmaze, etc.)
Better error handling and retry logic

License

[Add your license here]

Credits

Built to solve the debrid + public indexer problem that plagued Radarr/Sonarr users who don't want to pay for private trackers.

Support

Issues: Create an issue
Discussions: GitHub Discussions

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
tests		tests
.dockerignore		.dockerignore
.gitignore		.gitignore
Dockerfile		Dockerfile
Makefile		Makefile
README.md		README.md
TMDB_SETUP.md		TMDB_SETUP.md
docker-compose.yml		docker-compose.yml
main.py		main.py
pyproject.toml		pyproject.toml
requirements-dev.txt		requirements-dev.txt

natarelli22/pachelarr

Folders and files

Latest commit

History

Repository files navigation