Nhost

Nhost is the backend platform for agentic applications.

⭐ GitHub: Discord:

[Nhost @nhost]

Real ones remember.

What was the first programming book that started it all for you?

[Nhost @nhost]

Nhost is now running in Stockholm 🇸🇪

Lower latency for Nordic users. Pick it when you create a project.

... yes, the moose is a real design decision.

[Nhost @nhost]

Nhost is hiring an Infrastructure Engineer to own the multi-region cloud platform that every Nhost project runs on.

AWS, Kubernetes, Terraform, Go. Remote, lots of autonomy and ownership.

https://t.co/pAPu91zTC1

[Nhost @nhost]

🇦🇺 The Australia region is live!

Deploy closer to your users with Nhost’s new Sydney region.

Less latency, faster responses, happier users.

If your users are in Australia or APAC, your backend just moved a little closer. 🚀

[Nhost @nhost]

@VanshDhiwar9 @divssvash It’s your new favourite tech

[Nhost @nhost]

Sometimes, questioning your entire existence is a good thing.

[divyanshi @divssvash]

nhost up making me question my entire existence @nhost

[Nhost @nhost]

@divssvash Hey, would you mind sharing a couple of details: what OS and CLI version are you using?

[Nhost @nhost]

Edit rows straight from the table grid in the Nhost dashboard. Clone, edit, delete — no SQL.

[Nhost @nhost]

⌘K in the Nhost dashboard.

Search and jump to any page, setting, or project.

[Nhost @nhost]

@LangChain @useapolloio Apollo ditching supervisor graphs for dynamic skill selection is the right call. multi-agent DAG routing looks clean in a demo and gets messy in prod fast

[Nhost @nhost]

@AnthropicAI The judge meant to catch the sabotage shares the exact values that caused it, so your safety net and your failure mode are drawn from the same distribution.

[Nhost @nhost]

@DanielLockyer Definitely still a thing. We use GraphQL at Nhost because it's a great fit for working with complex, connected data, especially when you're building agentic apps.

[Nhost @nhost]

@nvidia @LangChain It's nice to see the open stack getting more love lately.

Good convo btw, worth the watch!

[Nhost @nhost]

@cursor_ai Excited to see what it actually unlocks beyond the coding benchmarks

[Nhost @nhost]

@AnthropicAI Finally an unlearning method where the knowledge actually leaves instead of getting suppressed and restored

[Nhost @nhost]

@LangChain Totally with you. Agents mostly assist humans today, but they're getting more and more autonomous. The infrastructure around them has to keep up.

[Nhost @nhost]

@langfuse huge for anyone doing multi-modal apps. great addition!

[Nhost @nhost]

@LangChain @bentannyhill Really like the shadow-production approach. Have you thought about running the agent against an isolated DB copy for the write-access problem instead of hand-mocking all the endpoints?

[Nhost @nhost]

@jerryjliu0 The quiet shift here is that retrieval starts looking more like an agent exploring a filesystem: search, grep, read, follow the thread.

That's much closer to how people actually research a case

[Nhost @nhost]

@GraphQL Good point on co-locating the fragment with the atomic component. The bit about a removed field making the dead UI obvious is the part people miss