Flox
@flox
flox init flox install <package> flox activate git add .flox/ git commit -m 'add repeatable dev environment' git push
Signing a person in and authenticating a pipeline are two different jobs. FloxHub now treats them that way. Today, we are announcing two new credential types for FloxHub: Service Account Tokens for automation, Personal Access Tokens for people. Full writeup: buff.ly/Qdhjrk9
Auto-activation is here for Flox environments. Walk into a directory with a .flox folder and the environment turns itself on. Nested environments and services come along too. Your environments become one less thing to think about. Read the full post here: buff.ly/8CYRuSbhttps...
One tool. Every language. Declare Python, Node, Go, Rust, or whatever your project needs in a single Flox environment. Flox activate, and the exact versions you pinned land first on PATH. Same on macOS and Linux, no containers. Change a version? That's a reviewable diff. buff.ly/de8Wb25
People keep asking how to set up their AI stack as code. So here's the pattern: one Flox file — packages, versions, services — that your whole team activates. Same stack on macOS, Linux, WSL2, CUDA or Metal. Link here: buff.ly/PVaRl8m
Getting the same environment twice was never easy. Now a coding agent installs dependencies and changes your system while you're not looking causing your laptop to drift further from prod. Flox CPO James Bayer gets into it on Temporal's Vibe Check: buff.ly/HWaXb38
Shai-Hulud doesn't break in. It gets installed, then reads whatever credentials are lying around and ships them off. A hermetic build keeps nothing lying around: no ambient creds, no network. It can't steal what it can't see. Link to full blog: buff.ly/Xw2Jie6
Couldn't make Planet Nix 2026? Or made it, but spent the whole time in the hallway track? Either way, the talk recordings are all on YouTube now. buff.ly/sG5nlLg
It trains on your MacBook then it hits Slurm and dies: wrong CUDA, missing libs, broken paths. Not another container. One Nix-pinned environment, identical from laptop to cluster to prod. We trained nanoGPT on Slurm to prove it: buff.ly/JJJd5yP
We're going live with Temporal on June 24. James Bayer shows how a reproducible Temporal + AI dev environment goes from a stack of manual setup steps to one command: `flox activate`. Then more on sandboxing, security & observability. 📆 Jun 24 · 11am PT · buff.ly/8xH3pQu
Planet Nix is back. April 1-2, 2027 in Pasadena. Two days of Nix in production — the talks, the workshops, and the people building with it. The CFP is open. If you've shipped something real with Nix, get on stage: buff.ly/sJ1qFRe
Resolve runs their developer environments on Flox: reproducible, the same on every machine. Read the story: buff.ly/vsoFqpL
Dev containers or Nix-style environments? False choice. Declare your runtime — language, libraries, build tools, CLIs — natively on the host with Flox, Nix, or Guix. Keep backing services in containers. The same split runs local → CI → prod. buff.ly/Y27BWgA
Flox is powering Resolve AI’s dev environments. Resolve triages production incidents for the world’s biggest companies. Their team now builds on reproducible Flox environments: one shared setup that can run locally, in CI, and in prod. Learn more here: buff.ly/m7m9TDo
New tools like DeepSeek TUI show why local AI tooling is heating up: Users want fast, terminal-native harnesses for local and cloud models on NVIDIA CUDA and Apple Silicon. We show how Flox and Nix make the stack—models, servers, CLIs, envs—easier to run, build, and share. buff.ly/RcQlnml
Fellow.ai hit a familiar scaling problem: “works on my machine” in local dev, CI/prod bugs, costly rework. With Flox, Fellow.ai got reproducible dev environments that work across laptops, CI, and prod—reducing onboarding friction, environment drift, and downstream risk. Full blog in the comments!
In financial services, today’s cutting-edge ML model becomes tomorrow’s baseline. Flox gives platform teams pinned, declarative GPU runtimes that travel from MacBooks to Slurm, CI, and Kubernetes—cutting runtime drift, image rebuild loops, and audit friction. Read the full article in the comments!
#PlanetNix 2026 throwback! Kelsey Hightower on discovering #Nix and realizing there was a fundamentally different way to think about software. buff.ly/9pzlu5c
fellow.ai used Flox to standardize reproducible environments across dev → CI → prod. The impact: ✔️ Faster engineer onboarding ✔️ Higher developer velocity ✔️ Fewer fire drills ✔️ Less rework ✔️ Reduced release risk Case study in comments.
We talk a lot about #AI right now, but this keeps coming up in conversations: determinism and reproducibility. If code is being generated by agents, and systems are getting more complex, you need to be able to answer a pretty simple question: what actually ran? Ron on this at PlanetNix 2026 👇
Security teams don’t need more heroic CVE triage. They need a better source of truth. Learn more about how reproducibility, traceability, and atomic response can reduce firefighting. buff.ly/a75igyP
New guide! A practical look at: - PyTorch runtimes that work across platforms - modular CUDA environments - build toolchains for R&D and MLOps - reproducible deployment patterns for AI teams buff.ly/M2OgkYX
A lot of people talk about frontier models. Less about how to run them reliably. New guide on running Gemma 4 and GLM 4.7 Flash locally with Flox using turnkey environments for Ollama and LM Studio. Reproducible local AI workflows across Mac, Linux, and Windows. buff.ly/0NeDdvS
In capital markets, speed matters. So do control and reproducibility. Learn how companies use Flox + Nix to reduce rebuild overhead, improve traceability, and promote or roll back by pinned reference. A practical look at reproducible AI infrastructure for regulated environments. buff.ly/h6wveJG
#AI increases the speed and volume of change across your entire software lifecycle. But your accountability stays the same. Flox provides deterministic foundations for probabilistic workloads, so velocity doesn't come at the cost of reliability, security, or compliance. buff.ly/VL4BuvB
We spoke with Farzan Karimi (Deputy CISO @ Moderna, SunsetCon) about what’s changing in #security right now: - Automated attack simulation is becoming standard - AI agents are taking over low-level pentesting + recon But the real shift? Agentic offense + agentic defense. buff.ly/t2fgYw7
Dependency resolution, version drift, conflicting toolchains, “works on my machine” → all of that burns tokens and breaks workflows. This is where Flox fits. 🟣 Cross-language package manager 🟣 Works across OS + CPU/GPU architectures 🟣 Deterministic, declarative environments buff.ly/RZtOBET
The real shift isn’t just “build faster with AI.” It’s zooming out and asking: how does all of this fit together? That’s where having a solid, reproducible foundation starts to make a big difference. Check out the full webinar here 👉 buff.ly/GWafnp6
