DJ Set Ripper

Extract tracklists from DJ sets and download each track individually.

⚠️ Legal Notice: This skill is intended for downloading music you have the right to access — purchases, free releases, creative commons, etc. Respect copyright laws in your jurisdiction. The author is not responsible for misuse.

Dependencies

Same as dj-mp3-sourcer (yt-dlp, ffmpeg/ffprobe, spotdl). No additional dependencies.

Workflow

1. Extract Page Content

Fetch the set URL and extract raw text (description, metadata, comments):

YouTube:
CODEBLOCK0

SoundCloud / Mixcloud:
Use web_fetch to grab the page content in markdown mode.

1001Tracklists:
Use web_fetch — this source has the most structured data. Prefer it when available.

2. Parse the Tracklist (LLM-Powered)

Feed the raw page content to the model with this prompt structure:

CODEBLOCK1

If parsing returns zero tracks, inform the user the tracklist couldn't be extracted and suggest:

• - Checking 1001Tracklists manually
• Pasting the tracklist directly

3. Download Each Track

For each parsed track (skipping any with artist AND title = "ID"):

1. 1. Use the dj-mp3-sourcer workflow: search sources in priority order, prefer extended mixes, download or surface purchase links
2. Use sessions_spawn to parallelize downloads (batch of 3-5 at a time to avoid rate limits)
3. Save files to: INLINECODE3

Set name is derived from the mix title (sanitized for filesystem).

4. Optionally Download the Full Mix

Ask the user if they also want the full mix downloaded. If yes:

CODEBLOCK2

5. Normalize Filenames

After all downloads complete (not per-batch — wait for every sub-agent to finish), run the normalization script once:

CODEBLOCK3

This fuzzy-matches each mp3 to a tracklist entry and renames to clean Artist - Title.mp3. Handles NA - prefixes, (Official Video) junk, wrong artist credits, label names, etc.

Critical: Run this in the parent agent after all batches return — do NOT rely on sub-agents to rename. The parsed tracklist is the source of truth for filenames.

6. Generate the Log File

Create ~/Downloads/{set-name}/{timestamp}.log with format:

CODEBLOCK4

Edge Cases

• - No tracklist in description — check 1001Tracklists via web_search: INLINECODE8
• "ID - ID" tracks — log as unidentified, don't attempt download
• Bootlegs / mashups — search anyway, but expect failures. log as not found with note
• B2B sets — multiple artists in set title, handle gracefully
• Duplicate tracks — deduplicate by artist+title before downloading
• Very long sets (50+ tracks) — batch in groups of 5, report progress as batches complete

Configuration