Your hub for all things AI Automation, Crypto & Infosec. Sharing insights, innovations, and best practices to shape a smarter, safer tech future.
Digital Nomad in USA!
Joined January 2022
- Tweets3.2K
- Following842
- Followers464
- Likes635
Yes! I agree!
This is 👇 what @calledtocnstrct and I talked about over mid week rambling on Spaces, on how agents need to be at kernel level. @bastidotnet validates it, and got it right and is working on implementing that concept.
#omarchy @omarchy
For me, an @omarchy phone must be agentic right down to the kernel level. No sandbox for the agents. It must be optimized for touch and can look different from the Omarchy desktop. It's still the real ARM Omarchy, but as a daily driver phone.
Update your Omarchy Uber TV guys. 0.4.1 update introduces tablet mode. If you havent tried out Omarchy Uber TV yet, you can find the link below
github.com/marcho78/uber-tv
Becareful with credentials guys
🚨 850+ compromised GitHub accounts. 13,500+ malicious branches.
Researchers uncovered 'Gheist,' an ongoing campaign using stolen tokens to plant GitHub Actions workflows that dump private repo secrets into logs.
Attackers download the logs, then delete the runs.
Details: thehackernews.com/2026/10/gh…
Devsec AI retweeted
I will actually send you $1 in 𝕏 Money…
Just reply with what you’re going to work on Monday to improve your life.
First 10 only, must have 𝕏 Money.
Could you have imagined few year ago you would be able to do this on a fly ? Contra clone built in Linux on top of Android using C?
Alright next up for Omarchy Uber TV is definitely tablet mode . When initially we built it it was meant for keyboard and soft remote . Now that I am testing on a tablet, I realize the need . So Omarchy Uber TV tablet mode coming up in next release.
Devsec AI retweeted
Infomarchy is ready for Omarchy 4.5
Infomarchy turns your desktop into a live info-desk overlayed to your background picture. Shows every AI agent you're running, what it's doing and your sub burn.
What Infomarchy needed to be Omarchy 4.5 ready:
O4.5 ships 22 usage collectors, one per agent. Infomarchy read the records alphabetically and capped them at 16, so ollama, openrouter, qwen, synthetic, vibe and zai got silently dropped. The cap is now 48, with a test that fails on the old one.
The new Claude collector dies with a KeyError when OMARCHY_PATH is unset, which would have left Claude limits stale with no warning. Infomarchy now passes it explicitly.
Qt 6.12 needed the palette singleton qualified as Commons.Color so theme colors resolve. Codex threads no longer double count after a rename.
Tested on 4.5.rc1 with Qt 6.12: 0 QML errors, the SUPER+D overlay works, and the Claude, Codex and Grok limit bars are live.
FREE and MIT:
github.com/nixfred/infomarch…
Which agents are on your desk right now?
Devsec AI retweeted
If you’re looking to connect to your Omarchy box from your Mac, give WayShare a try. Pretty painless process. 👇
github.com/wesleygrimes/ways…
This is nuts !
One TOS flag and your whole AI workflow is gone
@log4j says they were banned by OpenAI over local reverse-engineering work
Now picture that when everything you do runs through one API
My DGX Spark doesn't need anyone's permission to run, rent the frontier when you need it. Own a floor nobody can take away
Local wins!!
Comparing Opus 5.5 medium vs max mode is equivalent to one being highly detailed and has OCD, and the other as the brother with ADHD who wants to get shit done , but if it gets complex enough it will toss it aside or botch it.