Music Library

• Music Library
  • Features
  • Screenshots
    • Stats dashboard
    • Collection
    • Searching for a record to add
    • Record details
    • Record releases
    • Record chat
    • Artist details
    • Wishlist record details
    • Record sets
    • Scrobble rules
    • New scrobble rule
    • Universal search
  • Setup
  • Environment configuration
  • Running the application
  • Auditing Scrobble Data Quality
  • Deployment
  • CI
  • Architecture
  • Favicons

Features

Record management

• Add records from MusicBrainz, with optional override of specific pieces of data
• Cart-style multi-record import for adding several releases at once
• Manage a collection and a wishlist of records, with ways to quickly search and filter based on records' metadata (full-text search with structured query syntax: artist:X, album:X, genre:"Y", format:cd, type:album, purchase_year:2024)
• Browse record releases and select a collected release
• Dominant color extraction from cover art
• Generate a 120mm×120mm PDF tracklist from record and release data

AI chat

• AI-powered chat for records, artists, and the entire collection — with web search (OpenAI streaming via Responses API, entity-specific context for each chat type)

Artists

• Artist details with biography, discography, and similar artists
• Markdown notes on records and artists

Wishlist

• Wishlist with configurable online store templates for purchasing

Record sets

• Curate record sets (e.g. "best live albums") with drag-and-drop ordering

Search

• Universal search across collection, wishlist, artists, and record sets
• Similarity search via OpenAI embeddings (cosine-distance search with sqlite-vec)

Last.fm integration

• Display latest scrobbles, and where possible connect them with records in the collection or wishlist
• Browse and search the complete scrobble history
• Scrobble a record directly
• Store a local copy of the complete scrobble history, and setup scrobble rules to fix data quality issues (missing MusicBrainz IDs, artist/album remapping)
• Audit scrobble data quality via mix task: identify tracks with missing MusicBrainz IDs for artists and albums

Import & scanning

• Barcode scanning for quick imports (barcode → MusicBrainz lookup)

Stats & maintenance

• Stats dashboard with collection overview, top artists, top albums, and records on this day
• Daily email digest: "records on this day" with cover images and anniversary styling
• Admin maintenance dashboard: batch metadata/embedding refresh, database vacuum/optimize, Last.fm connection status

Infrastructure

• All data stored in SQLite databases for portability and ease of backup/restore

Screenshots

Stats dashboard

Collection

Searching for a record to add

Record details

Record releases

Record chat

Artist details

Wishlist record details

Record sets

Scrobble rules

New scrobble rule

Universal search

Setup

The project is managed and configured via mise-en-place:

• mise install will pull the correct Erlang, Elixir and Node.js versions
• mise run dev:setup will setup dependencies and database structure

Important

The project uses Fluxon UI, so it requires a valid set of credentials to pull the dependency. See the env section in mise.toml for the required environment variables.

It's recommended to use the git hooks included in the project. Install with:

mise generate git-pre-commit --write --task=dev:precommit

Environment configuration

Required environment variables for development are listed in mise.toml.

You can create a mise.local.toml with the required variables (sample values are included at the top of mise.toml).

For production, please see compose.yaml for a list of required variables.

Running the application

Start the Phoenix endpoint with mise run console (along with an attached IEx session).

Auditing Scrobble Data Quality

The application includes a Mix task to audit scrobbled tracks and identify data quality issues such as missing MusicBrainz IDs for artists and albums.

Running the Audit

# Audit all tracks
mix scrobble.audit

# Audit with detailed output including sample tracks
mix scrobble.audit --verbose

# Audit only artist issues
mix scrobble.audit --type artist

# Audit only album issues
mix scrobble.audit --type album

# Output as JSON for processing
mix scrobble.audit --format json

Understanding the Audit Report

The audit report shows:

• Total number of scrobbled tracks
• Artists with missing MusicBrainz IDs (grouped by artist name)
• Albums with missing MusicBrainz IDs (grouped by album title and artist)
• Track counts for each issue

Fixing Data Quality Issues

After identifying issues, you can:

1. Create Scrobble Rules: Navigate to the Scrobble Rules page in the web interface and add rules to map artist or album names to their correct MusicBrainz IDs.

2. Apply Rules: Use the "Apply Rules" button in the Scrobble Rules page to update existing tracks, or run in IEx:

   MusicLibrary.ScrobbleRules.apply_all_rules()
   

3. Re-audit: Run the audit again to verify the fixes worked.

The application also provides helper functions in the MusicLibrary.ScrobbleActivity context:

• count_tracks_missing_artist_musicbrainz_id/0
• count_tracks_missing_album_musicbrainz_id/0
• get_artists_missing_musicbrainz_id/1
• get_albums_missing_musicbrainz_id/1

Deployment

The application is deployed via Coolify, using a Docker Compose strategy.

CI

See the .github folder.

Architecture

See the docs folder.

Favicons

This favicon was generated using the following graphics from Twitter Twemoji:

• Graphics Title: 1f4bd.svg
• Graphics Author: Copyright 2020 Twitter, Inc and other contributors (https://github.com/twitter/twemoji)
• Graphics Source: https://github.com/twitter/twemoji/blob/master/assets/svg/1f4bd.svg
• Graphics License: CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)