Sherlock tops GitHub's daily trending as 9 repos add 7,172 stars
These 9 repos give Western developers a week's worth of starred patterns for username lookup, bare-metal GPU Kubernetes, and AI browser control, without wading through the whole trending feed.
GitHub's daily trending for 2026-11-09 pulled together 9 open-source projects that added 7,172 stars in a single day. Sherlock, a Python CLI that searches for usernames across social networks, led with 2,373 new stars, bringing it to 103,537 total. Python and TypeScript dominated the list's languages.
Beyond sherlock, the roundup includes Nvidia's NVKE, a management plane for bare-metal GPU Kubernetes clusters, currently flagged as Preview and not meant for production; OpenMausBot, a local-first alternative to Grok Bot that runs Claude, Codex, or Grok agents in sidebars; and Agent Browser, a Rust CLI that gives AI agents a snapshot-based control over a live browser. The oddest entry is billion-context, a TypeScript plugin that compresses conversation contexts into hierarchical summaries, claiming to drop context window needs to ~100K and token usage to about one-fifth.
The roundup itself is AI-generated, as the footer notes, so the list is as much algorithmic curation as a manual pick, meaning star counts are real but the selection may reflect bot-like filtering.
Agent tooling dominates the top of the list: OpenMausBot, agent-browser, claude-quickstarts, and skills all target AI agent workflows, marking this as the most active open-source niche this cycle.
billion-context's fivefold token reduction claim is a headline number worth verifying; the plugin works by rewriting streams and injecting compression tools, so real savings depend on how well hierarchical summaries compress actual conversations.
NVKE's honest Preview status, with explicit limits on management-cluster upgrades and backup recovery, is a useful counterweight to the usual production-ready hype around Kubernetes tooling.