GitHub Trending: Today — Spotube, ReVanced, GOWA, lm-evaluation-harness, public-api-lists
Today's list is unusually consumer-facing: a desktop-and-mobile music client, an Android patch manager, and a WhatsApp API sit beside model evaluation and a public-API catalog.
In today's GitHub Trending: Spotube (open-source music streaming), ReVanced Manager (Android patch management), GOWA (multi-account WhatsApp REST API), and lm-evaluation-harness (language-model testing).
The mix is a reminder that developer attention can cluster around tools that make closed platforms more usable, even when the community signal is thinner than the star count.
The roster
| Repo | What it does | Daily growth |
|---|---|---|
| KRTirtho/spotube | Open-source music streaming app | +30 stars |
| ReVanced/revanced-manager | Android patch manager | +20 stars |
| aldinokemal/go-whatsapp-web-multidevice | Multi-account WhatsApp REST API | +23 stars |
| EleutherAI/lm-evaluation-harness | Few-shot language-model evaluation framework | +20 stars |
| public-api-lists/public-api-lists | Searchable catalog of free public APIs | +23 stars |
KRTirtho/spotube — a music client with an audience beyond developers
Spotube is for listeners who want an open-source music app across desktop and mobile rather than another developer-facing dashboard. Its place beside infrastructure tools is the day’s oddity.
The research pass found a little HN discussion and no Lobsters result, while GitHub shows active issues. That is not a verdict on the app; it is a useful distinction between a repository people install and one that prompts a broad technical debate.
EleutherAI/lm-evaluation-harness — benchmarks remain plumbing
For teams comparing language models, lm-evaluation-harness supplies a framework for few-shot evaluation. It is the least consumer-visible entry in the roster and the one closest to the AI work hidden behind model-launch charts.
Its current discussion signals are similarly modest: historical HN references and active GitHub issues, but no fresh cross-community conversation. That absence matters because evaluation tooling is often essential precisely when it is not glamorous.
The day’s practical rule is to separate adoption from discourse: a project can be useful enough to surge without generating a new argument about how it works.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
Anthropic says agent teams need social engineering, not just smarter models
A swarm can split a task, but that does not make it a team. Anthropic's new research argues that durable cooperation needs environments, incentives, and institutions built for agents.