Skip to main content
AI/MLjmagly

tag-collection

Apply metadata tags, embed artwork, and organize media files with consistent naming

Stars
141
Source
jmagly/aiwg
Updated
2026-05-31
Slug
jmagly--aiwg--tag-collection
View on GitHubRaw SKILL.md

// install — copy + paste into any project

mkdir -p .claude/skills && curl -fsSL https://raw.githubusercontent.com/jmagly/aiwg/HEAD/agentic/code/frameworks/media-curator/skills/tag-collection/SKILL.md -o .claude/skills/tag-collection.md

Drops the SKILL.md into .claude/skills/tag-collection.md. Works with Claude Code, Cursor, and any agent that loads SKILL.md files from .claude/skills/.

/tag-collection

Apply metadata tags, embed artwork, and organize media files with consistent naming conventions.

Purpose

Automate the metadata curation workflow for audio and video collections:

  1. Scan files in a directory
  2. Look up canonical metadata from MusicBrainz
  3. Apply tags using opustags (Opus) or ffmpeg (MP4/MP3)
  4. Embed cover artwork
  5. Rename files to follow naming conventions
  6. Organize into proper directory structures

Parameters

Required

  • <collection_path>: Path to directory containing media files to process

Optional

  • --artist <name>: Filter to specific artist (speeds up processing)
  • --dry-run: Show what would be done without making changes
  • --artwork-dir <path>: Path to canonical artwork directory (default: ./artwork)
  • --force: Overwrite existing tags and files without prompting
  • --skip-artwork: Skip artwork embedding (metadata only)
  • --skip-rename: Apply tags but don't rename files
  • --skip-organize: Apply tags and rename but don't move files

Workflow

1. Scan Files

# Find all supported media files
opus_files=$(find "$collection_path" -type f -name "*.opus")
mp4_files=$(find "$collection_path" -type f -name "*.mp4")
mp3_files=$(find "$collection_path" -type f -name "*.mp3")

2. Parse Filenames

Extract artist, title, album from existing filenames using common patterns:

  • Artist - Title.ext
  • Artist - Album - Track# - Title.ext
  • Track# - Title.ext (when artist is provided via --artist)
# Example parsing
filename="Joni Mitchell - Both Sides Now.opus"

artist=$(echo "$filename" | sed -E 's/^([^-]+) - .*/\1/' | xargs)
title=$(echo "$filename" | sed -E 's/^[^-]+ - ([^.]+)\..*/\1/' | xargs)

3. Lookup Metadata

Query MusicBrainz API for canonical metadata:

# URL-encode artist and title
artist_encoded=$(echo "$artist" | jq -sRr @uri)
title_encoded=$(echo "$title" | jq -sRr @uri)

# Query API
mbdata=$(curl -s "https://musicbrainz.org/ws/2/recording/?query=artist:${artist_encoded}%20AND%20recording:${title_encoded}&fmt=json" \
  -H "User-Agent: MediaCurator/1.0")

# Extract metadata
album=$(echo "$mbdata" | jq -r '.recordings[0].releases[0].title')
year=$(echo "$mbdata" | jq -r '.recordings[0].releases[0].date[:4]')
tracknumber=$(echo "$mbdata" | jq -r '.recordings[0].releases[0].media[0].track-offset + 1')
genre=$(echo "$mbdata" | jq -r '.recordings[0].releases[0].release-group.primary-type')

# Rate limit: 1 request per second
sleep 1

4. Fetch Artwork

Download cover artwork from MusicBrainz Cover Art Archive:

# Get release MBID
mbid=$(echo "$mbdata" | jq -r '.recordings[0].releases[0].id')

# Download front cover
curl -s "https://coverartarchive.org/release/${mbid}/front" -o "/tmp/cover-${mbid}.jpg"

# Fallback to fanart.tv if CAA unavailable
if [ ! -f "/tmp/cover-${mbid}.jpg" ]; then
  curl -s "https://webservice.fanart.tv/v3/music/${mbid}?api_key=${FANART_API_KEY}" | \
    jq -r '.albums."'"$mbid"'".albumcover[0].url' | \
    xargs curl -s -o "/tmp/cover-${mbid}.jpg"
fi

5. Apply Tags

Use appropriate tool based on file format:

Opus Files (opustags):

opustags "$file" \
  --set "TITLE=$title" \
  --set "ARTIST=$artist" \
  --set "ALBUM=$album" \
  --set "ALBUMARTIST=$artist" \
  --set "TRACKNUMBER=$tracknumber" \
  --set "DATE=$year" \
  --set "GENRE=$genre" \
  --set-cover "/tmp/cover-${mbid}.jpg" \
  -o "${file}.tagged"

mv "${file}.tagged" "$file"

MP4 Files (ffmpeg):

ffmpeg -i "$file" -i "/tmp/cover-${mbid}.jpg" \
  -map 0 -map 1 -c copy \
  -disposition:v:1 attached_pic \
  -metadata title="$title" \
  -metadata artist="$artist" \
  -metadata album="$album" \
  -metadata date="$year" \
  -metadata genre="$genre" \
  "${file}.tagged.mp4"

mv "${file}.tagged.mp4" "$file"

MP3 Files (ffmpeg):

ffmpeg -i "$file" -i "/tmp/cover-${mbid}.jpg" \
  -map 0 -map 1 -c copy \
  -id3v2_version 3 \
  -metadata title="$title" \
  -metadata artist="$artist" \
  -metadata album="$album" \
  -metadata date="$year" \
  -metadata genre="$genre" \
  -metadata:s:v title="Album cover" \
  -metadata:s:v comment="Cover (front)" \
  "${file}.tagged.mp3"

mv "${file}.tagged.mp3" "$file"

6. Rename Files

Apply naming convention based on file type:

Audio Files: {Artist}/{Album}/{Track#} - {Title}.{ext}

# Construct new filename
new_filename=$(printf "%02d - %s.opus" "$tracknumber" "$title")
new_path="${artist}/${album}/${new_filename}"

# Show rename operation
echo "RENAME: $file -> $new_path"

# Execute if not dry-run
if [ "$dry_run" != "true" ]; then
  mkdir -p "$(dirname "$new_path")"
  mv "$file" "$new_path"
fi

Video Files: {Artist}/{Collection}/{Title} [{Quality}].{ext}

# Detect video quality
quality=$(ffprobe -v error -select_streams v:0 \
  -show_entries stream=height -of default=noprint_wrappers=1:nokey=1 \
  "$file")

quality_label="${quality}p"

# Construct new filename
new_filename="${title} [${quality_label}].mp4"
new_path="${artist}/${collection}/${new_filename}"

echo "RENAME: $file -> $new_path"

if [ "$dry_run" != "true" ]; then
  mkdir -p "$(dirname "$new_path")"
  mv "$file" "$new_path"
fi

7. Organize Files

Move files into directory structure if not already done during rename:

# Ensure artist directory exists
mkdir -p "$artist"

# Move album directories
if [ -d "$album" ]; then
  mv "$album" "$artist/"
fi

8. Report Changes

Generate summary of operations:

echo "=== Tagging Summary ==="
echo "Files processed: $file_count"
echo "Tags updated: $tags_updated"
echo "Artwork embedded: $artwork_count"
echo "Files renamed: $renamed_count"
echo "Files moved: $moved_count"

if [ "$dry_run" = "true" ]; then
  echo ""
  echo "DRY RUN: No changes were made. Run without --dry-run to apply changes."
fi

Dry-Run Mode

Use --dry-run to preview operations without making changes:

/tag-collection ~/Music/Joni\ Mitchell --dry-run

Output shows what would be done:

SCAN: Found 47 Opus files, 12 MP4 files
LOOKUP: Joni Mitchell - Both Sides Now
  -> Album: Clouds (1969)
  -> Track: 12/14
ARTWORK: https://coverartarchive.org/release/a1b2c3d4.../front
TAG: Set TITLE, ARTIST, ALBUM, TRACKNUMBER, DATE, GENRE
EMBED: cover-a1b2c3d4.jpg
RENAME: Joni Mitchell - Both Sides Now.opus -> Joni Mitchell/Clouds/12 - Both Sides Now.opus

DRY RUN: No changes were made. Run without --dry-run to apply changes.

Examples

Basic Usage

Tag all files in a directory:

/tag-collection ~/Music/Unsorted

Filter by Artist

Process only files for a specific artist:

/tag-collection ~/Music/Unsorted --artist "Joni Mitchell"

Preview Changes

See what would be done without making changes:

/tag-collection ~/Music/Unsorted --dry-run

Skip Artwork

Apply metadata tags but skip artwork embedding (faster):

/tag-collection ~/Music/Unsorted --skip-artwork

Custom Artwork Directory

Use artwork from a specific directory:

/tag-collection ~/Music/Unsorted --artwork-dir ~/Media/artwork

Metadata Only (No Rename/Move)

Update tags but leave files in place with original names:

/tag-collection ~/Music/Unsorted --skip-rename --skip-organize

Force Overwrite

Overwrite existing tags without prompting:

/tag-collection ~/Music/Joni\ Mitchell --force

Interactive Prompts

When not using --force, the command prompts before overwriting:

File already has complete metadata:
  Title: Both Sides Now
  Artist: Joni Mitchell
  Album: Clouds
  Year: 1969

Overwrite existing tags? [y/N]:

Error Handling

MusicBrainz Lookup Failure

If MusicBrainz API returns no results:

WARNING: No MusicBrainz match for "Unknown Artist - Unknown Track"
SKIP: Will not tag this file

Manual intervention required - add to skip list or provide metadata manually.

Artwork Download Failure

If cover art is unavailable:

WARNING: No cover art found for album "Rare Demos"
CONTINUE: Tags will be applied without artwork

File is still tagged, just without embedded artwork.

Rate Limit Exceeded

If MusicBrainz rate limit is exceeded:

ERROR: MusicBrainz rate limit exceeded (503 Service Unavailable)
PAUSE: Waiting 60 seconds before retry...

Command automatically waits and retries.

File Permission Errors

If file cannot be written:

ERROR: Permission denied writing to "file.opus"
SKIP: File will not be modified

Check file permissions and ownership.

Quality Checks

After completion, verify:

  1. Tags Applied: Check random files with opustags file.opus or ffprobe -show_format file.mp4
  2. Artwork Embedded: Open files in media player, verify artwork displays
  3. Filenames Correct: Check naming convention is followed
  4. Organization: Verify directory structure matches expectations
  5. No Duplicates: Ensure no files were duplicated during moves

Performance

Typical processing speeds:

  • Metadata lookup: ~1 second per file (MusicBrainz rate limit)
  • Tagging (opustags): <0.1 seconds per file
  • Tagging (ffmpeg): 1-3 seconds per file (re-encodes)
  • Artwork download: 1-2 seconds per album
  • Total: ~100 files in 2-3 minutes

For large collections (1000+ files), expect 20-30 minutes.

Troubleshooting

"opustags: command not found"

Install opustags:

# macOS
brew install opustags

# Ubuntu/Debian
sudo apt install opustags

# Arch
sudo pacman -S opustags

"ffmpeg: command not found"

Install ffmpeg:

# macOS
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

# Arch
sudo pacman -S ffmpeg

MusicBrainz Timeout

Increase timeout in curl command:

curl --max-time 30 -s "https://musicbrainz.org/ws/2/recording/..."

Artwork Not Embedding

Verify image format and size:

# Check image file
file cover.jpg

# Resize if too large
convert cover.jpg -resize 1200x1200 cover-resized.jpg

See Also

  • /organize-media - Organize files without metadata changes
  • /extract-artwork - Extract artwork from existing files
  • Skill: @metadata-tagging
  • Skill: @cover-art-embedding
  • Agent: @Metadata Curator

References

  • @$AIWG_ROOT/agentic/code/addons/aiwg-utils/rules/human-authorization.md — Seek explicit authorization before bulk-overwriting metadata tags on existing files
  • @$AIWG_ROOT/agentic/code/frameworks/media-curator/skills/metadata-tagging/SKILL.md — Metadata tagging patterns used by tag-collection
  • @$AIWG_ROOT/agentic/code/frameworks/media-curator/skills/cover-art-embedding/SKILL.md — Cover art embedding patterns used by tag-collection
  • @$AIWG_ROOT/agentic/code/frameworks/media-curator/skills/curate/SKILL.md — Orchestration skill that invokes tag-collection as a curation phase