I’ve curated the ultimate must-follow list of top X accounts covering global politics, AI leaders, macro policy, business thinking & alternative cognition.
Build your high-efficiency information pipeline in one place.
Full list 👇
I built a bridge between Claude Skills and Codex because I kept running into the same problem:
Good context was getting trapped inside one ecosystem.
Claude and Codex are both powerful, but switching between them often means rebuilding instructions, workflows, and agent behavior from scratch.
So I built:
codex-claude-bridge-skill
What it does:
• translates Claude-oriented context into Codex-friendly workflows
• reduces duplicated setup work
• preserves reusable agent behavior
• makes moving between ecosystems less painful
Repo:
https://t.co/YS1Ekr60dr
My take:
The future isn't one AI tool replacing another.
It's interoperability.
Developers shouldn't have to migrate their brains every time they switch agents.
Elixir v1.20 isn't just another release.
It's a signal.
For years, one of the biggest questions around Elixir was:
"Great concurrency... but where are the types?"
v1.20 starts changing that.
Key improvements:
• broader type inference
• earlier bug detection
• faster compilation
• smarter recompilation
• stronger tooling foundations
The interesting part isn't "Elixir got types."
It's that more correctness is happening automatically.
Developers write code.
The compiler increasingly understands intent.
That's a very different future from manually adding annotations everywhere.
Release notes:
https://t.co/GDE1BK4UIz
My take:
The best developer tools don't ask humans to do more work.
They quietly remove work humans shouldn't be doing.