We're sharing/showcasing best of @github projects/repos. Follow to stay in loop. Promoting Open-Source Contributions. UNOFFICIAL, but followed by github
San Francisco, CA
Joined October 2020
- Tweets13.6K
- Following66
- Followers342K
- Likes23.9K
GitHub Projects Community retweeted
Experiential is an agentic AI gateway.
Use hosted, BYOK, and local models through one OpenAI-compatible API, then turn real AI traces into simulations for automated evals and training.
For a limited time, Experiential is also offering 1000+ models free through its hosted catalog at platform.xplabs.ai
GitHub Projects Community retweeted
What will be the output of the following Python code?
x = 1 << 4
print(x & (x - 1))
52%0
16%1
15%8
16%15
310 votes • Final results GitHub Projects Community retweeted
📘 Linear Algebra Done Right — Free PDF
A great resource for anyone learning:
• Linear algebra
• Vector spaces
• Linear transformations
• Eigenvalues & eigenvectors
• Abstract algebra concepts
📄 404 pages
If you’re learning mathematics, machine learning, AI, or data science, this is a valuable book to add to your collection.
💬 Comment “ALGEBRA” and I’ll share the link.
🔁 Repost/Share so others can access this free resource.
GitHub Projects Community retweeted
Introducing Impulse Purchase Cooldown™
If checkout detects an impulse buy, your payment is locked for 10 minutes so you can “think about it.”
GitHub Projects Community retweeted
Kimi-K3 just got 62% compressed to run locally with 79% accuracy intact.
Kimi K3 can now be run locally! ✨
The 1-bit model retains ~78.9% accuracy after we shrunk it from 1.56TB to 594GB (-62% size).
Run on a Mac Studio + 128GB RAM device.
Kimi K3 is the strongest open model to date.
Guide: unsloth.ai/docs/models/kimi-…
GGUF: huggingface.co/unsloth/Kimi-…
GitHub Projects Community retweeted
Mercury Agent v1.1.13 is here.
Most AI agents are powerful.
Few actually remember.
Mercury Agent’s Second Brain gives agents persistent relational memory, allowing them to remember people, projects, decisions, and context across sessions without repeatedly consuming tokens to rebuild state.
Built for long-term workflows, not one-off conversations.
Highlights:
• Second Brain relational memory
• Scoped Permissions with approval-first execution
• Unified Skill Management
• Token Saver for smarter context usage
• 31 built-in tools
• Long-running autonomous workflows
• Millisecond memory retrieval
• Bring your own OpenAI-compatible models and endpoints
Faster recall. Lower token usage. More context retained.
Because the future isn’t bigger context windows.
It’s agents that remember.
Your agent. Your rules.
GitHub Projects Community retweeted
Why use multiple tools when one engine does it all?
iii is the easiest way to compose, extend, and observe every service in your stack in real time.
Every backend starts as a project before the first line of business logic. Queues, cron, HTTP, state, observability, agents, and sandboxes each usually bring their own integration story. iii collapses that into one live system surface.
iii worker add
GitHub Projects Community retweeted
Mercury agent is growing, glad to see how much efforts are here behind this open source project.
Mercury agent has a new home.
mercury.cosmicstack.org → mercuryagent.sh
Same powerful personal AI agent.
Same soul-driven, token-efficient, always-on experience.
Now on a cleaner domain built for what Mercury is becoming.
GitHub Projects Community retweeted
GitHub says the recent unauthorized access incident started from a poisoned VS Code extension installed on an employee device.
Not a zero-day.
Not some movie-style breach.
Just one trusted developer tool inside the supply chain.
GitHub Projects Community retweeted
Mercury agent introduces a serious memory architecture for agents.
A dual-memory layer with conscious + subconscious context, paired with agentic Kanban/workspace that can plan, recall, and execute across workflows.
This is how agents move from reactive to genuinely smart.
GitHub Projects Community retweeted
OmniGet is one of those tools you don’t realize you need until you try it.
Paste a link → it pulls media from YouTube, Twitter, TikTok, Reddit, even Udemy
Native desktop app. No ads. No tracking. Just downloads that work.
Simple idea, executed properly.
GitHub Projects Community retweeted
Most multimodal stacks still work like this:
Image encoder → LLM → generator → formatter
Every handoff adds latency, complexity, and information loss.
SenseNova-U1 by @SenseTime_AI is pushing a cleaner path:
One native unified system for understanding, reasoning, and generation.
That’s the architectural shift worth watching.
GitHub Projects Community retweeted
OpenClaw showed agents can act.
Hermes showed they can remember.
Mercury solves the next problem:
An agent that stays.
Always there when needed.
Memory that compounds.
Identity you own.
Permissioned execution.
Presence that persists.
Soul-driven. Token-efficient. Always on.
GitHub Projects Community retweeted
Same prompt. Three video models. One of them stopped looking like AI.
We ran Seedance 2.0 against Veo 3.1 and Wan 2.7 on three prompts designed to stress-test video generation: atmosphere and reflections, sustained physics in a Japanese forge, and a chaotic Tokyo fish market at 5am.
Seedance held up in ways we didn't expect.
Seedance 2.0 is now live on @gmi_cloud's console. They're running a Seedance 2.0 Comparison Challenge, with $100 in credits awarded to the five most creative entries.
Try it: console.gmicloud.ai