for anyone wanting to migrate off railway for cloud VMs, sandboxes, CI at boat.dev we're up to 23x less expensive at scale our VMs have 99.95% uptime we offer the highest perf CPU in the world we have a great CLI for agents to migrate you off in a day

seems obvious to everyone around me that sandboxes are on their way out. we hear too many stories about reliability, capacity, broken sandboxes that need manual intervention -- plus they are way too expensive for what they provide. the future for agents? workspaces. your agent doesn't need a computer. it needs a place to store it's context, and the *tools* to manipulate that context each turn. we're in the very early stages of the AI era. the simplest version of this pattern is just handing each agent a time-multiplexed VM, since that's the only thing that we have at hand right now. over the next 12 months, you're going to see a rapid shift away from this model and towards what i call the "magic compute box". rather than giving an agent a VM, you give it a set of tools (could be "bash", "python", "js", "read_file", "write_file") -- and you send those commands to a magic service in the cloud that manages to run them for you -- using the cheapest technology available (just-bash, wasm, or even Linux if really required) for each command invcation. your agent doesn't want a computer that it can set a desktop background for, it wants a place where it can fire off an unlimited number of tool calls and have it update the context correctly.

you either die a sandbox provider, or live long enough to become a neocloud sandboxes are a commodity offering. they’re all more/less the same. there is a set of 10 or so features you need. the user does not care that it boots 100ms faster. move into inference, RL, and serverless functions offerings. capture all of the compute. big compute providers have economies of scale they benefit from through these use cases. If you are only offering one use-case, you’ll have difficult keeping pace. similarly big enterprise co can manage their own fleet of sandboxes and capture cost savings via reserved instances, surplus capacity, and even spot instances. if you’re into sandboxes, pivot into general compute

here

Unpopular opinion: You need coding knowledge to vibe code properly.

I'm moving away from my local dev setup Makes zero sense to me now

good afternoon today we tracked down and obliterated 7,013 fraudulent accounts this amounts to an estimated $400,000 a month in savings which helps OpenCode Go become more sustainable thank you for your attention to this matter

I very much agree with the original take, see zerops.io For perhaps slightly different reasons, though. It mainly has to do with environment parity. Since the dawn of time, if you wanted your developers to produce good code, you wanted to give them an environment that is as close to production as possible. The more complex the project and the bigger the team, the more this made sense. But the reality is that this is almost never the case. For local development people spin things up with Docker Compose, some might use remote environments like GitHub Codespaces / Gitpod (which absolutely do not match production), stage would perhaps be on some self-managed cluster, production then on AWS or some PaaS. Even before the agent craziness we set out to build a platform where you'd have environment parity throughout the whole development lifecycle. It's simply good development practice that shouldn't only be available to teams with a dedicated Ops crew able to put a complex K8s setup together. Each project gets a private VXLAN network with dedicated L3/L7 balancers, statistics and logs containers, the same container primitive (a full Linux OS, so usable both as a remote development container and as a production container), the same managed services. Whether something is a remote development environment for a developer or production is just a matter of what you named the project. Guess who else absolutely loves this setup? Coding agents. Whatever they produce running remotely on Zerops is tested against real infrastructure, whatever they produce is production ready end to end, because they essentially can work against production. Whatever they produce can be spun up as an identical environment for stage, or for another colleague, human or human with agent. Google, in one of their recent papers, argues that the harness makes up 90% of the intelligence/capabilities, and having created a couple of projects on Zerops with agents, I very much agree - proper infrastructure is the last missing piece of this harness onion, and what agents are able to produce when all the layers are in place is wild.

Thank you, everyone, for participating! We are overwhelmed with your projects. 👏

My coding & deployment & provisioning dreams come true. Well done @fxckdev_ and team. Love it! zerops.io/zcp #zerops #development #agents #ai #mcp #cli

my entire 2026 stack runs on cloudflare and I pay $5/month for it - workers (compute) - d1 (sql db) + better auth - kv (cache) - r2 (object storage) - queues (background jobs) - hyperdrive (db acceleration) - email routing (sending) - + waf & ddos built-in

I’d estimate we’re ~6 months from most devs moving their code agents off of their laptops

Two prompts with Fable in @Lovable: 1/ "build a copy of Spore space-age part, where you can fly your spaceship between stars, zoom into planets and upgrade your ship etc." 2/ "Make it multiplayer."

We just launched Sites into Codex! Software creation was always about more than writing code. Sites in Codex fundamentally gives the power of end-to-end software creation to every user, no matter their technical fluency. These Sites are fully deployed to a URL, private to workspaces, come with authentication, can have static files, and can store dynamic data in databases. It is in preview for business and enterprise teams and will be rolling out to all workspaces over the next day. Give it a try by typing @ Sites into Codex and ask it to build anything! This project took a massive amount of effort across hundreds of people at OpenAI - proud that we were able to get this out and excited to see what you all build with it!

Railway's built in SSH console 🤝 Hermes Agent by @NousResearch Video tutorial dropping soon