GitHub listed only nine projects this Monday morning, and five of them were featured here in the past week. Today's four picks are two first appearances, MuJoCo after its 4.0 release and Anthropic's Claude API quickstarts, and two earlier picks that have materially changed: Anthropic's skills repository, whose Office document skills moved to Apache 2.0 on 6 November, and OpenMausBot, which has shipped sixteen releases since we covered it on 30 October. Curation order reflects practical interest, not GitHub rank. Trending is not a release date or an endorsement.
Picks are listed in curation order, not GitHub’s rank; each card shows the rank GitHub displayed when the page was checked. Stars measure attention, not security, quality or fitness for use, and no project here was installed or run. “Stars today” is GitHub’s own displayed figure for its daily window, not an audited rolling 24-hour count.
What it does. MuJoCo (Multi-Joint dynamics with Contact) is Google DeepMind's general-purpose physics engine for robotics, biomechanics and machine learning. It has a C API, Python bindings and a Unity plug-in, and simulates articulated bodies touching their environment.
Caveat. 4.0 is a major version with several breaking changes: new tendon side-site semantics, the cable plugin and composite element removed, Python 3.10 support dropped. The release notes warn that upgrading the pip package in place needs an uninstall or forced reinstall first.
Attention, not endorsement
GitHub showed
“41 stars today” as displayed for GitHub’s daily window; not an audited rolling 24-hour count
What it does. Anthropic's public collection of Agent Skills: folders of instructions, scripts and resources that Claude loads for a task, plus the Agent Skills specification and a template. It holds 19 skills, from design and testing to the docx, pdf, pptx and xlsx document skills.
Caveat. GitHub still detects no repository-wide licence: each skill carries its own LICENSE.txt, and doc-coauthoring has none. The README calls the skills demonstration and educational material whose behaviour may differ from Claude's. Skills include scripts an agent runs, so review them like code.
Attention, not endorsement
GitHub showed
“792 stars today” as displayed for GitHub’s daily window; not an audited rolling 24-hour count
What it does. A desktop chat app for macOS, Windows and Ubuntu where each AI bot is a contact with its own model, computer and connected apps. Bots run on the claude, codex or grok command-line tools already signed in on your machine.
Caveat. A meeting bot records other people, so participants' consent is needed. Outside a MausBot Cloud plan, transcription needs your own ElevenLabs, Groq or OpenAI key, or a local Whisper model. Cloud computers come from the paid third-party Boat service after a trial, and near-daily releases mean frequent change.
Attention, not endorsement
GitHub showed
“1,439 stars today” as displayed for GitHub’s daily window; not an audited rolling 24-hour count
What it does. Anthropic's starter projects for building on the Claude API: a customer-support agent, a financial data analyst, computer-use and browser-use demos, minimal browser and computer toolset drivers, an autonomous coding agent, and a growing set of Managed Agents examples for Slack, Linear, Sentry and more.
Caveat. Every quickstart needs a Claude API key and paid usage. The README says the toolset examples are not production code, and the macOS computer-use reference controls a real desktop, which it advises running in a VM.
Attention, not endorsement
GitHub showed
“32 stars today” as displayed for GitHub’s daily window; not an audited rolling 24-hour count
Every repository GitHub listed, in GitHub’s order, with the figures it displayed when the page was checked (): 4 picked today, 5 covered in an earlier digest, 0 not covered. For projects covered before, growth compares the total stars at our first pick with today’s displayed total.
TypeScript · 460 stars today · 5,399 stars in total
A context-compression plugin for month-long single sessions (billions of tokens) ,small context windows (a 100K context is enough) and token savings (5x fewer tokens) .上下文压缩插件,月级长会话会话, 兼顾小窗口(100k上下文足矣)省token(省5倍token) . billion-context is all you need
Covered before — first picked 8 Nov 2026 at 4,731 stars; now 5,399: +668 stars (+14.1%) in 1 day.
Following Our Picks
Every project picked in an earlier digest, biggest growth first. “Then” is the total at our first pick’s metadata check; “now” is GitHub’s API star count checked . Growth measures attention, not quality or safety.
GitHub's daily listing held only 9 rows at capture, fewer than on any day of the past week (13 to 18). The raw page was checked: it contains exactly nine listing rows, so the short list is GitHub's, not a parsing gap.
All 9 listing entries were inspected. Exact public GitHub API metadata and recent releases were checked at about 07:02 UTC for the four picks and for vercel-labs/agent-browser; README files of the four picks were read at about 07:02:30 UTC; the anthropics/skills commit history since 31 October, the changed LICENSE.txt files before and after commit 2c5958e, and each skill folder's LICENSE.txt were read at about 07:04 UTC. Every checked repository was active and unarchived.
Five entries featured in the past seven days were left out: sherlock-project/sherlock, dsx-ai-factory/nvke, meta/astryx and ranxianglei/billion-context (all picked yesterday, 8 November) and vercel-labs/agent-browser (6 November; latest release still v0.41.2 of 3 November).
Two picks are repeats outside the seven-day window, each for a primary-sourced change: anthropics/skills (first picked 31 October) for its 6 November licence change, and milind-soni/OpenMausBot (first picked 30 October) for releases v0.1.135 to v0.1.150, including a meeting-notes bot in v0.1.147.
The preceding seven calendar dates, 2-8 November, were checked for local history; all seven editions exist.
The official listing was captured at 08:01:38 Europe/Paris. Literal stars-today labels are GitHub's displayed period figures; API total stars and README evidence were checked separately at the times shown per card. They are not an independently audited rolling 24-hour growth measurement.
Feature descriptions come from each project's own README and release notes and are the maintainers' claims. No candidate code, installer, binary, plugin, skill or model was installed, downloaded or run, and none of this is a security audit or endorsement.