We are excited to announce the release of OSS Insight's organization analysis 🎉
You can gain more flexible insights into the performance of open-source organizations, regardless of your project repository's organizational structure.
https://t.co/ocXTAxHBlh
🦀 The Rust Shift in AI
7 Rust agent infra repos in 60 days. zeroclaw 30K ⭐. agent-browser 28K ⭐.
Python for models. Rust for runtimes. The AI stack is splitting — just like web infra did a decade ago.
https://t.co/NMxYLvM2M0
#Rust#AI#GitHub#OpenSource@zeroclawlabs
🏢 The Zero-Human Company Wave
50K+ ⭐ for paperclip in 38 days. AI agents don't assist — they ARE the company.
Devs aren't voting on architecture. They're voting on the metaphor.
https://t.co/aEpVfgfCGL
#AI#GitHub#OpenSource#AIAgents
📊 DESIGN.md — GitHub's fastest design standard
35K ⭐ in 10 days. Markdown is becoming the protocol layer between humans and AI agents.
AGENTS.md → CLAUDE.md → DESIGN.md → SKILL.md
The .md Protocol Layer is real 👇
https://t.co/P7TYuw5Rg5
#AI#GitHub#OpenSource #AIAgents
Thanks to recent "#opensource" event, I’ve strengthened my #agentteam tool which help me run OSSInsight— essentially, #harnessengineering
❓ Intent recognition
📷 Task decomposition
📷 Scoring
📷 Progress tracking
🕐 Progress tracking
☕️ Delivery result
Take a look👇
AI Agents own the chart: #openclaw 23k★, everything-claude-code 12k★, agency-agents 12k★
@karpathy 's #autoresearch hit 12k★ in weeks
Claude ecosystem on fire — skills, learn-claude-code, get-shit-done all Top 20
#MiroFish, #RuView, #deerflow rising across different tracks
Skill repos are growing 1.5× faster than agent runtimes on GitHub.
4 skill packs vs 4 runtime engines—distribution layer wins the star race, with 35× less issue pressure.
We dug into the data ↓
https://t.co/GoboIEdcqy
#AI#OpenSource#GitHub
6 repos, 130k+ stars, 90 days. A new GitHub pattern: agent-native CLIs.
Software is being re-wrapped so AI agents can call it—just like REST APIs enabled mobile 15 yrs ago.
Data ↓
https://t.co/Aeik7T1r7u
#AI#GitHub#OpenSource#Agents
Running @OSSInsight solo? No problem !
#AgentTeam ? NO! I built an Agent *Nation* ��
With 148 AI agents. One human. Each with its own role and expertise. Drop a message, the right agents
respond with structured proposals.
This is what #HarnessEngineering looks like 👇👇👇
@MSitarzewski I wanted one system to manage #OpenClaw, #ClaudeCode, #Codex, and other #AI agents together.
Running @OSSInsight taught me: the bottleneck isn’t model quality. It’s coordination.
Who wakes up?
Who responds?
Who owns the work?
Who closes the loop?
That’s why I built ClawRoom.
I am here to #connect with people who are interested in:
- Build in Public
- Full Stack
- Startup
- Tech
- AI
- SaaS
Let's 🤝 and connect to grow together.
50K ⭐ for a config file.
We analyzed the data behind @garrytan's gstack, @mattpocock's & @slavingia's skills repos.
Star-to-fork ratios of 8-15x reveal: people star to learn how someone thinks, not to use the code.
https://t.co/RXwr4XKHz2
#AI#GitHub#OpenSource
4 teams rewrote OpenClaw from scratch. 4 languages. 116K ⭐ in 8 weeks.
🐍 nanobot: 5,551 lines of auditable Python
🦀 ZeroClaw: $10 hardware, 5MB RAM
🟣 NanoClaw: 8,778 forks — they're not using it, they're building on it
https://t.co/uQyahAOcI6
#OpenSource#AI#GitHub
54K stars in 19 days.
karpathy/autoresearch doesn't automate research — it turns research into search.
You sleep. It runs 100 experiments overnight. Keeps what works, reverts what doesn't.
https://t.co/UV7v1EHQMX
GitHub's trending is a black box. We built our own.
🔥 https://t.co/FICmvYFW2b
10.5B GitHub events. Filter by language. Full analytics per repo.
7/8 top trending repos right now? AI agents.
Blog: https://t.co/yFT9oe72Ot