Bryan FRIMIN retweeted
I just ordered my chromium mirror repo. Let’s see how they ship 66G repo on a CD
(CD limit is ~700Mb)
We heard you. And we agree.
In light of recent developments in physical media, GitHub is proud to announce that you can now obtain your public repo on CD-ROM.
Keep it. Lend it to friends. Pass it on to your children.
Your code is physically yours, forever. Until you lose it, let's be real.
Order yours today.
gh.io/cd
Claude Managed Agents is now available on Blaxel.
Run your agent loop on @ClaudeAI managed infra, while running tool execution on secure, persistent sandboxes.
-> 25ms suspend/resume. Shared persistent filesystem. Controllable network.
xitter.cf/ClaudeDevs/status/2065…
Claude Managed Agents can operate in a sandbox you control, on your own infrastructure or with any provider you choose.
Today we added new guides for @blaxelAI, @e2b, @googlecloud, @namespacelabs, and @superserve_ai, so you can choose the best fit for your use case.
Bryan FRIMIN retweeted
Proudly announcing PgDog's $5.5M seed round from @BasisSet, @ycombinator, @pioneer_fund, and others.
Running Postgres in prod is hard. PgDog is a modern connection pooler, load balancer and database sharder. It can be deployed anywhere and works with all Postgres hosts, incl. RDS Postgres, Aurora, Cloud SQL, Supabase, Neon, etc.
Some cool numbers: 2M+ queries per second in prod, 20TB+ sharded, 50+ open source contributors.
1 year down; here's to 10 more.
Bryan FRIMIN retweeted
Today, we're officially introducing the new Vybe.
We started as "Lovable for internal apps." We spent the last year building the infrastructure: permissions, integrations, databases, workflows. Everything companies need to actually trust software inside their business.
Then, OpenClaw came out and started showing what was possible with agents for personal use cases. Everything clicked for us.
The infrastructure we'd built wasn't just useful for humans anymore: it was exactly what agents needed to operate inside real companies and teams.
What happens when agents can work 24/7, with the right context, the right permissions, and access to your company's tools?
A founder becomes a one-person army. A small team operates like a much bigger one.
That's the mission we fell in love with at @vybe_build. Our agents have names and roles. They have their own email, Slack, phone number. They prep meetings, follow up with leads, run ops, update CRMs, build internal tools.
Vybe runs on Vybe.
We're onboarding 10 teams a week to run AI-native companies.
Bryan FRIMIN retweeted
My advice to founders in 2026: spend tokens, not headcount.
Record everything. Make your company queryable. Build self-improving loops.
Before long, AI won’t just help you operate your company. It will make it self improving.
Don't think AI adoption, think AI transformation.
This is the biggest shift in how startups get built since cloud computing.
Bryan FRIMIN retweeted
Biggest red flag I hear from early stage founders?
“We’re working on self-serve onboarding”
Your product currently requires you to talk to every new customer?
That’s a good thing.
Bryan FRIMIN retweeted
If you’re a French YC founder (current batch or alumni) and want to meet other French YC founders in SF 🇫🇷
We’re hosting a curated founder event this Wednesday, May 27.
If you’re interested, let me know.
Bryan FRIMIN retweeted
The essential engineering cheatsheet of 2026:
agent → while loop
subagent → nested while loop
agent harness → the rest of the code
cloud agent → all the above, on EC2
Blaxel is officially a first-class sandbox provider in the new version of @OpenAI Agents SDK.
Your agents now have shell, file I/O, and live preview URLs, all running in isolated cloud sandboxes on Blaxel infra.
xitter.cf/OpenAIDevs/status/2044…
OpenAI Developers@OpenAIDevsApr 15🇺🇸
United StatesConnected via WebAccount-level information, not a live location or per-post device.
Replying to @OpenAIDevs
Bring your own environment.
The Agents SDK now supports sandbox execution with providers including @Cloudflare, @Vercel, @modal, @e2b, and more.
Keep files, credentials, and execution state in your environment while passing approved context to the model.
Bryan FRIMIN retweeted
We're bringing the advisor strategy to the Claude Platform.
Pair Opus as an advisor with Sonnet or Haiku as an executor, and get near Opus-level intelligence in your agents at a fraction of the cost.
Bryan FRIMIN retweeted
Whoops....
github.com/apache/httpd/comm…
Bryan FRIMIN retweeted
Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software.
It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans.
anthropic.com/glasswing
Bryan FRIMIN retweeted
Orange a commencé depuis 00h
l'extinction de son réseau 2G en France. Le GSM, lancé en 1992, tire sa révérence après 34 ans de service. Un thread sur pourquoi c'est plus qu'un simple changement technique 🧵👇
Bryan FRIMIN retweeted
Yeah there’s a line between “we stand by founders no matter what” and “we won’t tolerate bullshit and lying to your customers”
You can legitimately be both. You just need boundaries and a back bone.
Bryan FRIMIN retweeted
If you use GitHub (especially if you pay for it!!) consider doing this *immediately*
Settings -> Privacy -> Disallow GitHub to train their models on your code.
GitHub opted *everyone* into training. No matter if you pay for the service (like I do). WTH
github.com/settings/copilot/…





























