# Changelog of Genius Scraper — Lyrics, Credits, Samples & Annotations (`haketa/genius-scraper`) Actor

- **URL**: https://apify.com/haketa/genius-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/haketa/genius-scraper.md

## Changelog

### 0.2.0 — 2026-11-01

- Verified a fresh **400-song, multi-artist run** (Drake, Taylor Swift, Kendrick Lamar, Eminem, Kanye West) with **0 duplicates** and **100% coverage** on the core identity fields (`songId`, `title`, `artist`, `url`) plus `releaseDate`, `producers`, `writers`, `primaryGenre`, `tags`, `language`, `explicit`, `pageviews`, `songArtImageUrl`, `description`, `annotationCount`, and `songRelationships`.
- Confirmed the signature fields work at scale: **lyrics on 96%** of songs (median ~3,400 chars), **annotations on 94%** (10,900+ annotations total), and the samples/interpolations graph (`songRelationships`) returning **14,000+ relationships** across the run. Run finished in ~16 minutes over residential proxies with no value bugs (no `[object Object]`, correct types, all absolute URLs).
- Confirmed artist-catalog **pagination** (popularity-sorted, `next_page` cursor) and in-run **de-duplication** by Genius song ID both behave correctly, and that `maxItems` / `maxSongsPerArtist` cap the run exactly as documented.
- Refreshed the README to the gold standard: replaced the output example with a real captured record, documented honest per-field coverage (noting that `featuredArtists`, streaming links, and `album` are present-when-available), corrected the `hot` field description (omitted when not hot), and refreshed the scale/pagination numbers.
- Added this changelog.
