maoty

maoty is a small Astro site that surfaces albums worth tracking. The homepage is built from generated JSON data and is intended to be refreshed from external music sources.

What is in this repo

• Astro frontend in src/pages/index.astro
• Generated album dataset in src/data/album-list.json
• Album aggregation script in scripts/build_album_data.py
• Last.fm library scraper in scripts/scrape_lastfm_library.py

Local development

Install dependencies:

bun install

Start the Astro dev server:

bun run dev

Build the production site:

bun run build

Preview the production build:

bun run preview

Data workflow

The site reads album entries from src/data/album-list.json.

To refresh that file:

bun run albums

scripts/build_album_data.py currently does all of the following:

• reads the Last.fm export at output/mameli_mixtape_first50_artists_with_tags.json
• derives a tag profile summary and writes output/mameli_mixtape_tags_browse.md
• opens Album of the Year pages, collects album metadata, and merges the result into src/data/album-list.json

External prerequisites for bun run albums

The album build is not fully self-contained. It expects:

• python3
• playwright-cli available on PATH
• a usable Playwright session/profile for Album of the Year scraping
• the Apple Music helper script at $HOME/.codex/skills/apple-music-album-linker/scripts/find_apple_music_album.py
• the Last.fm export file at output/mameli_mixtape_first50_artists_with_tags.json

If any of those are missing, scripts/build_album_data.py will fail early.

Create the aoty Playwright session

scripts/build_album_data.py is hard-coded to use:

• session name: aoty
• profile directory: .playwright/aoty-profile

Create that session once from the repo root:

playwright-cli -s=aoty open about:blank --headed --persistent --profile .playwright/aoty-profile

What this does:

• -s=aoty gives the browser session the exact name the script expects
• --persistent stores cookies and browser storage on disk instead of keeping them only in memory
• --profile .playwright/aoty-profile keeps that persistent profile inside this repo
• --headed opens a visible browser so you can complete any login or Cloudflare challenge manually

Recommended first-run flow:

1. Run the command above.
2. In the opened browser, visit Album of the Year and complete any login or anti-bot challenge if needed.
3. Leave the browser open and run bun run albums, or close it and reuse the saved profile later.

Useful session commands:

playwright-cli list
playwright-cli -s=aoty open about:blank --headed --persistent --profile .playwright/aoty-profile
playwright-cli -s=aoty close

Notes:

• The script checks whether the aoty session is already open and will attach to it if available.
• If the session is closed, the saved profile at .playwright/aoty-profile is what preserves cookies between runs.
• If Album of the Year starts returning Just a moment..., reopen the same aoty session headed and refresh manually before rerunning the build.

Last.fm scraper

Use the standalone scraper to create or refresh the Last.fm export:

python3 scripts/scrape_lastfm_library.py \
  "https://www.last.fm/user/<user>/library/artists" \
  --output-base output/lastfm_library_artists_with_tags

Useful flags:

• --workers controls concurrent artist tag requests
• --max-pages limits how many library pages are scraped
• --max-artists caps the total number of artists written

The scraper writes both CSV and JSON outputs.