If you are a VSCode hater (like me), you can achieve the same thing using the open-source, Microsoft-built, official devcontainer CLI [0], which is exactly what VSCode uses under the hood. You don't get automatic port forwarding with this tool, which includes SSH Agent forwarding, so you'll likely want to pair this with installing SSH into your devcontainers, or using some other tool that does the job. I built a tool [1] to do the port forwarding part, which I'd like to be more magic like VSCode's but haven't put the effort into doing, since it works for my purposes.
Using these, I can SSH into my devserver, use the devcontainer CLI to open a tmux session with iTerm2 Tmux integration, and it feels just like a local window. Combine that with Neovide and remote Neovim over the same channel, and I have a native-feeling dev environment running on a remote devcontainer in a nearby data center.
Not loving that the devcontainers CLI is node. The appeal of devcontainers is to limit the amount of code I have to trust on my host. Give me a small static binary I can deploy to new systems.
I ssh into my server, then use the devcontainer cli tool to open a shell in the dev container. Found that easier than installing ssh in every container, especially when there's lots of them. I made a simple command line tool for helping the process, so I don't have to type out any long commands. If I run it it will list the running dev containers and I give them each a number to I don't have to use the full ID.
Yeah, this works. Two drawbacks: you have to rebuild the container when the port forwarding configuration changes, and the port forwards have to be unique on the entire machine.
I like the CLI for commands, but would rather work with an agent chat interface for vibing, which unfortunately I’m becoming increasingly reliant on, even though it’s gotten me into trouble with larger projects.
I have many environments locally, some dependent on others, and some require local databases. I use containers in production, but not always locally.
It’s almost a hellscape situation for trying to setup a fully isolated dev environment. I might be able have have some future generation of Claude convert my existing dev setup to something isolated with k8s. But, I don’t have the power to run such an environment locally, with all of the apps and services. I’d need to set my dev environments up in the cloud, but I don’t have enough money to pay for that, and it wouldn’t even work in some circumstances, because some things must run locally.
So, instead I run Cursor AI with various MCP servers, I approve tool calls, I don’t have things isolated, and I know this will eventually bite me.
> which unfortunately I’m becoming increasingly reliant on, even though it’s gotten me into trouble with larger projects
You could… not do this. You know it causes you problems, you consider it unfortunate that you’re becoming more reliant on it, and yet for some reason you’re choosing to continue anyway. Why are you doing that?
In a similar situation here: it's "unfortunate" in one sense but "extremely fortunate" in another sense.
It's unfortunate if you consider your primary craft to be writing code, as those skills can atrophy. It's obviously important to review and understand all material produced by these tools.
It's fortunate if your primary organizational role is "fix problems, communicate technically, and make new things" because of the sheer power of these tools when applied properly.
I've sunk many hours into attempting to "claude-ify" a complex set of applications and services and (IMO) that's a very useful activity.
Why? So that I can use ai-tooling more effectively? Yes.
Why also? So that services are more decoupled, more testable, more aligned with good development principles.
It's difficult to inject these workflows in a useful way across multiple levels (code/devops/org) but when it works it's worth it.
My key takeaway was something along the lines of: "if an agent can't understand and work with your codebase, you've got an onboarding problem" (we do).
Disclaimer: mostly standard web tech across Java/Scala/React - aggressively complex k8s layer
Just tried it... and it says it's client-only, but when I try to run the SSH provider using my Rapberry Pi it says it needs `docker` to be installed on it :/. So it's not just client-only, is it?
No, client-only means Devpod is contrasting itself to Coder, which requires a central server to orchestrate the containers it creates. Devpod does not. It does still have dependencies, though, among them Docker.
Mate, if you need to install something on a server how the hell can it be called client-only??? eMacs is an example of client-only perhaps, with TRAMP as it works purely on top of SSH , though you could argue it still isn’t client-only as it requires the SSH server to be enabled… so I thought the SSH connector for this would also work the same way but it also requires a full freaking Docker install not just SSH. This is totally misleading.
Oh the gitlab link is really really really cool. I can't express in words how happy I am since I know about iroh and I always wanted to create such a tool myself too but never could figure out iroh.
Now the way I did port forwarding was really wild and really complicated and I think that it might demand a blog post if I am being honest, but yes this project is absolutely godsend. Thanks!!
It's proprietary software masquerading as open-source, for one, and intentionally fosters dependence on its vendor for the plugins that support the most popular languages as well as key features (remote editing, WSL support, integration with the mainline plugin repos), for one.
They do indeed use a different implementation of these features than the one that ships with VSCode, though they're not the original authors of it. They used to host their version separately here: https://github.com/anysphere/open-remote-ssh
But I guess they've just rolled it into the Cursor repo? Idk, because Cursor itself is proprietary.
There's a similar extension for WSL integration, though I'm not 100% sure if Cursor's implementation of this feature is based on it: https://github.com/jeanp413/open-remote-wsl
You can sideload proprietary VSCode extensions to add them to VSCodium or some downstream fork, but you may also have to patch the extensions themselves because some of them refuse to run if they aren't being used with Microsoft's VSCode: https://github.com/OliverKeefe/vscode-extensions-in-vscodium
Microsoft injects proprietary spyware into the official builds that you can't see from the source code. You can get unofficial builds of just the source code from the VsCodium project, but Microsoft blocks your ability to install a lot of extensions if you don't use the official build.
It’s clumsy as hell. It’s neither a good editor, which focus on being fast when doing things with text, nor a good IDE, which is more about the tooling and getting a project done. It’s trying to be both at the same time. And the result is bad at both.
I don’t know, others might be better out of the box. But once you start adding features they all become a mess. I rather just deal with a known mess and not waste time with tools.
For small files and quick edits i use neovim with no customisation at all. Back when I first started using vim I had a pretty long config and some plugins installed and so on. But now that I no longer try to use it as an IDE, neovim is perfect for small edits.
For development work I use JetBrains suite of tools. Very minimal amount of customisation there also, only a couple of extra plugins installed by me and a couple of changes in the settings. It works well enough that I don’t have any desire to spend time customising it a whole bunch.
Don’t get me wrong, I loved neovim and other tools but tbh I just nano for simple edits now. AI tab complete is just too good. Basically it’s barebones or VSCode (or its variants). I could probably use vanilla neovim instead of nano but I've not used vim all that much recently to be in that "flow".
Your point is good, I don't mean to undermine it, but because I work in this area I feel compelled to point out that depending on the type of Swift you're writing using NeoVim to write Swift is actually quite pleasant: https://www.swift.org/documentation/articles/zero-to-swift-n...
I'd waste time with tools. I have my vim config, my git config, my emacs config, my bash config,... that I copy over to any development environment I have. And every now and then, I go into the manual or other people config to check if there's any new trick I can apply.
Comfort is something invaluable when doing a task. I can accept discomfort when I'm in an unfamiliar environment, but not when it's going to be something I will be in for days.
Ultimately, they're about as trite as reasons to hate Vim or Emacs. It boils down to preference, and I'm outlining a path to using another editor and getting access to some of the same features.
Each time I try to use it, I find myself distracted by the impulse to fix the bugs that I notice. There's always this moment when I find a comment from myself, several years earlier, in an issue that's still open because it's unclear whether it's a VSCode problem or a plugin problem, and it's like: "oh, here I am again, doing this instead of working".
That's not to say it's especially buggy, it's just so large, there are a lot of places for bugs to hide.
I get distracted by the impulse to contribute to other editors also, but those are contributions that I'm happy about. They don't always result in bugs fixed or features added, sometimes somebody says "no" but at least it resolves in some way. With VSCode it feels like I'm going in circles.
I personally feel some inner sense of dissonance when I use software written in Electron when alternatives are available. Software targeted to the technical audience has no valid reason not to aim higher in terms of performance and fundamentals.
I do not hate it when others are using it (none of my business). But I do not like to use bloatware. Text editor that eats RAM (even excluding LSP) like 3d editing software.
Just so you know, the devcontainer VSCode extensions are considered proprietary by Microsoft and therefore not allowed to be installed in VSCodium. There are some open-source container plugins, but they are all a bit janky for the container use-case in my experience.
I have actually been working on something similar but instead of Docker, I am using Apple’s builtin container[1] support to run shell commands and code with real OS-level isolation. It’s fast (spawns in milliseconds) and integrates nicely with things like Claude Code and the Gemini CLI. I open sourced it as CodeRunner[2]. Would love to hear what people think or chat about how it compares.
I have enjoyed running Claude Code in a container. The biggest advantage for me isn't security though, it's how easy it becomes to spin off agents to work in the background.
I use a simple script that copies my working directory into a container, prompts Claude Code, and then either saves a planning document locally, or opens a pull request in GitHub for me to review.
I quite like this because it makes starting agents when I have ideas really frictionless, and then I can easily discard the results and try again if it went off the rails. The Max plan makes me not worry about the cost of this whole process as well.
I also set up containers initially out of caution, but honestly, I've never seen Claude Code do anything that git couldn't easily reverse, so I'm not that worried about that any more.
Yeah the point of security is missed. Given how easy it is, I doubt I'm ever going to run Claude code or any other CLI that uses remote AI servers without a devcontainer again
> I've never seen Claude Code do anything that git couldn't easily reverse, so I'm not that worried about that any more.
I had one agent that installed something with pip and used it to rewrite my git history. It used the wrong flag and what was left was one file. I am glad I caught it before auto-commit and push triggered. It was one of my hobby projects, so not a huge deal. I can't imagine the fallout of this happening in a work environment.
I have, but I just found that a bit tedious to manage manually. Maybe another version of this script would just be to automatically create a worktree, run a prompt in that, and then have the script manage the results, and clean up the worktree at the end.
One thing that I’ve found works well is to permanently have a few numbered branches checked out as worktrees, keep them up to date with main, and then when you want to do a task in parallel you just pick an existing branch rather than creating and deleting worktrees every time.
This lets you persist local databases, IDE project settings and so on for each worktree too.
They can, but I find it harder to review if you are trying to get Claude Code to do anything non-trivial. I've seen some people make multiple git worktrees so they can have two instances of Claude Code working independently, but I found that to be tedious to manage.
Not the OP, but I have this fish alias. I just run `claude-docker` in my project root (where I would otherwise run `claude`). There's nothing to it. You can tweak it:
alias claude-docker='docker run --mount type=bind,source=/Users/hboon/.claude,target=/home/node/.claude --mount type=bind,source=/Users/hboon/.claude.json,target=/home/node/.claude.json --mount type=bind,source=.,target=/workspace/project --workdir /workspace/project -it cc /usr/bin/fish'
I'm afraid it is not that shareable as it contains a lot of dependency management and setup that is specific to my setup and projects. But it's not too complicated. You could probably re-create your own version fairly quickly.
It's just a Dockerfile that installs dependencies, an entrypoint that invokes claude, and some wrapper scripts that handle creating the container, passing through the prompt, and managing the results (e.g., output markdown to a directory, or open a PR). For input/output from the container I mount directories within the Docker container to local directories under ~/.call_claude/. I create a directory under there, clone the source code to it, output the prompt to a text file, and then grab results from that directory after Claude Code has finished running.
It's just not that simple. If your scripts contain 20 things hard-coded to your setup and projects, it is not quick nor easy to remove all of those and make something configurable and easily adaptable to other people's setups and projects.
Adapting the scripts to your specific setup is pretty much the entirety of the work in setting this up.
Yes please, more containers-first thinking for all things AI. The fact that this has been largely absent from most discussions and demos regarding agents and vibe-coding seems like a big red flag to me. I mean if we're going to YOLO the code, perhaps we should at least pay some attention to the dev / test harness?
It's worth thinking about reproducibility even if you're not concerned about security. We would not have to argue so much about the effectiveness of something like vibe-coding if everyone could simply share process/tooling without tons of custom setup. And even if you're not worried about filesystem sandboxing and leaking secrets like ssh keys, even if versioning with git is enough of an undo button for you.. who wants to clutter their whole system with a ton of new language stacks and/or weird dependencies for every small experiment? Most experiments with the new hotness are going to fail, so we should expect to do a lot..
Containers bring their own set of problems, there are some examples brought up in this thread, mainly around communication with the host OS.
I‘d argue the reproducible parts of vibe coding (agentic engineering) setups are just text files. Many people use a mix of web apps (AI studio), Mac apps (Wispr Flow), and other UI tools (repo prompt) in their workflow which can’t be put in a container anyway - well, reasonably at least.
If you want security, containers won’t get you that far. You’ll need to use a VM.
But if you give Claude Code access to your GitHub repo, what else is there worth protecting, that’s not protected by keychain & sudo?
All development is text files, that is missing the point. The development environment is a system, and a pretty complicated one too. It matters where the files are, what's in them, and how they interact. Things change together instead of staying isolated, you add more pieces over time, and even more things need to change together. Anyone who likes text-files more than click-to-configure UIs for tools, will probably like containers more than text files for systems, and for all the same reasons.
Your choices to reproduce complex systems are basically to 1) deny that complexity exists and accept any corresponding limitations in your working environment, 2) follow some error-prone multistep processes to reproduce existing system setup manually, 3) commit to centralizing all development on external cloud platforms, or 4) do something else to bundle together a setup that's repeatable.
I'm strongly in favor of (4) here, and while I'd probably disagree that it requires VMs instead of docker, the goal of repeatable processes is so important that hey, whatever works. It sounds like you're in camp 1 or 2?
> The development environment is a system, and a pretty complicated one too.
It's also partly personal. I also like to use isolated environments, mostly because I tinker with different technologies, and over times, it becomes a mess on my machine.
But I still like my git config, my editor and other tooling that I have to assist in my tasks. And it's not worth it to replicate it over to other people's setup. And neither I want their things.
> But I still like my git config, my editor and other tooling that I have to assist in my tasks. And it's not worth it to replicate it over to other people's setup.
This isn't about stopping you from customizing your development environment. It's about making your custom development environment automatically useful on your desktop or your laptop, or committing a project-specific starter environment to version control so that it doesn't take everyone on your team days to onboard.
> And neither I want their things.
Why not? This isn't about something like git aliases or editor keyboard shortcuts. It's about tooling pipelines. If someone's claiming enhanced productivity I don't want to watch their 1 hour youtube video, pause it, and painstakingly recreate their setup to check if the claims are true.
There's a reason projects grow Tiltfiles, compose files, and similar. In the limit, agents are basically just a collection of microservices that just happen to operate on code. How do you expect to do that sort of thing without tools?
And you don't need to stop there. Claude Code also has great open alternatives you can run in any environment you like, with any model, and under economic arrangements you define. See https://oss-ai-swe.org/ for a few. My current favs are OpenHands and Codex CLI.
Yes and I really hope more models consider pricing like Claude Max. There's also the option of choosing local models but I haven't tried that yet. Claude is just too good to spend time elsewhere (at the moment).
Containers do seem to work really well for agents. It lets them run in parallel! Giving them sudo also lets them do interesting things you wouldn’t want them doing on your desktop, eg. installing tools or debugging with tcpdump. (I am working on agent-in-a-container: sketch.dev)
Letting Claude Code loose on my home folder feels bad yes. Would using a separate user account solve it? I.e. keep git repos in for example /home/claude and run claude as that user? Or am I missing something?
This is exactly why I’m using chatGPT codex over Claude Code (even though I suspect CC might be better). Each chat runs in its own cloud based container. There’s no risk of exposing secrets/deleting files/breaking environment. I can run 3 in parallel while I continue to code locally. It works in a branch and integrates via PR. It’s a great DevEx.
Someone please build a nice macOS app using hypervisor framework that lets me do this locally :)
Honest question: why do people prefer developing code inside a docker? I get the benefits of docker as a deployment unit, but wouldn’t configuring a dev container and using it a hassle nonetheless, compared to not doing them at all?
Personally I don’t trust running npm/pip/gradle/etc in my own machine. I feel more secure running them in containers or VMs. Some IDEs like jetbrains’ support remote development (it’s not perfect but better than risking my own machine)
That's a different thing. You're talking about running the software under development in a container. I think the commenter was asking about running the dev tooling in a container, which I find a massive pain.
No, that’s what they’re talking about - running the tooling in a container. Dev tool chains are a pain in the ass to keep consistent in even small teams. Devcontainers solve that.
You could include those as part of the tooling. I have been experimenting with including mise as part of the image and then layer on the extra tools within there. Put all of those steps into the build so it is automatic.
If you’re like me and want to maintain a clean dev machine. I don’t want my machine to get in my way when I need to get work done. That means I don’t want some X messing things up somewhere that I didn’t know about. Older I get my memory isn’t as sharp and things like this matter.
I use VMs myself. I use Proxmox and have it setup so that I can spin up a fresh VM in around 10-20 seconds when needed. I also like that I can take snapshots.
I really want to limit the amount of software I am installing on my main system due to possible security issues.
I have recently moved to a VM only development workflow, but still feeling some growing pains as I figure out a good setup. I run my IDE in the guest, which is kind of crummy - responsiveness is worse + the VMs are a lot heavier than a headless code setup would be. I have thought about using a Flatpak IDE with filesystem access disabled as a potential middle ground solution so that the guests could be minimal images connected by SSH.
If you are using Claude Code on macOS running it in a devcontainer has a few upsides like fewer cli tool call failures, meaning less waiting time and less context window spam.
But there are some things you lose as well @ ergonomics:
- "done" notifications are not working by default (might be possible trigger and send to host using hooks feature)
- more difficult to give playwright a session (need to get cookies into playwright in the container)
- can't easily paste screenshots into the devcontainer (Claude Code has to take a screenshot or you have to put it into a mounted dir)
and of course the usual advantages of using containers still apply (dependency management, reproducibility and so on).
I still wonder though if nix would allow me to remove the container overhead (they do noticeably tax the macbook battery).
Claude Code using grep is often failing (also in devcontainer but less often). Many command line tools on macOS work subtly different than the ones on linux.
This setup is also compatible with GitHub CodeSpaces, which I think is an exceptionally nice development experience if you are on something like a Chromebook, and gives you isolation from your local network. Also lets you share your app running in dev-mode with anyone for feedback, without having to poke holes in your network or whatever. Also also, you can use the Cline extension in this environment if you still want the cursor-style copilotey experience.
From what I can gather, this is part of the idea/appeal behind something like https://phoenix.new/. The AI Agents have total control over the environment and can use all of the associated system tools as a result, without putting your own machine at risk. Once things are kinda working, you can pull it out and work locally.
Is this level of fear typical or reasonable? If so, why doesn’t Anthropic / AI code gen providers offer this type of service? Hard to believe Anthropic is not secure in some sense — like what if Claude Code is already inside some container-like thing?
Is it actually true that Claude cannot bust out of the container?
Just a month ago, an AI coding agent deleted all the files on someone's computer and there was a little discussion of it here on HN. Support's response was basically "yeah, this happens sometimes".
HN thread (flagged, probably because it was a link to some crappy website that restates things from social media with no substantive content of its own): https://news.ycombinator.com/item?id=44262383
It's worth noting that the default settings of Cursor do prevent this by asking you to confirm every command that is run. And when you get tired of that 5 minutes in and switch to auto-approving there is still protection against files outside the work directory being deleted. The story above is about someone who disabled all the safeguards because they were inconvenient, then bad things happened
It is a good example of "bad things can happen", but when talking about whether we need additional safeguards the lessons are less clear. And while I'm not as familiar with the safeguards of Claude Code I'm assured it also has some by default
If the cointainer route (with those types of privileges) is being suggested from a security point of view, then might as well run these processes in a VM and call it a day :/
I haven't found that to be the case. I have used cc within an container and on the host machine and it has been fine. Any command that could cause changes to your system you MUST approve when using it in agent mode.
You either have the option of approving each command manually, or you can let it run commands autonomously. If you let it run any commands then you have the risk of it doing something stupid (mainly deleting files).
You also have MCP tools running on your machine, which might have security issues.
I have personally never seen claude (or actually any AI agent) do anything that could not be fixed with git. I run 24/7 in full permissions bypass mode and hardly think about it.
Anyone with more than one toolbox knows that fear isn't required. Containers are about more than security, including stuff like organization and portability.
> If so, why doesn’t Anthropic / AI code gen providers offer this type of service?
Well perhaps I'm too much the cynic, but I'm sure you can imagine why a lack of portability and reproducibility are things that are pretty good for vendors. A lack of transparency also puts the zealots for "100x!", and vendors, and many other people in a natural conspiracy together, and while it benefits them to drum up FOMO it makes everyone else burn time/money trying to figure out how much of the hype is real. People who are new to the industry get leverage by claiming all existing knowledge does not matter, workers who are experienced but looking to pivot into a new specialization in a tough job market benefit from making unverifiable claims, vendors make a quick buck while businesses buy-to-try and forget to cancel the contract, etc etc.
> Is it actually true that Claude cannot bust out of the container?
Escaping containers is something a lot of people in operations and security have spent a lot of time thinking about long before agents and AI. Container escape is possible and deadly serious, but not in this domain really, I mean all your banks and utility providers are probably using Kubernetes so compared to that who cares about maybe leaking source/destroying data on local dev machines or platforms trying to facilitate low-code apps? AI does change things slightly because people will run Ollama/MCP/IDEs on the host, and that's arguably some new surface area to worry about. Sharing sockets and files for inter-agent comms is going to be routine even if everyone says it's bad practice. But of course you could containerize those things too, add a queue, containerize unit-tests, etc
Of course. Also with regular customer projects. Even without AI--but of course having an idiot be able to execute commands on your PC makes the risk higher.
> If so, why doesn’t Anthropic / AI code gen providers offer this type of service?
Why? Separate the concerns. Isolation is a concern depending on my own risk appetite. I do not want stuff to decide on my behalf what's inside the container and what's outside. That said, they do have devcontainer support (like the article says).
>Hard to believe Anthropic is not secure in some sense — like what if Claude Code is already inside some container-like thing?
It's a node program. It does ask you about every command it's gonna execute before it does it, though.
>Is it actually true that Claude cannot bust out of the container?
There are (sporadic) container escape exploits--but it's much harder than not having a container.
You can also use a qemu vm. Good luck escaping that.
Or an extra user account--I'm thinking of doing that next.
I don't get these posts. I'm using claude --dangerously-skip-permissions all day and haven't had a single issue. In my experience it doesn't just randomly start erasing your hard drives. Also having a proper CLAUDE.md probably helps.
You ignore safety features unless you have already experienced disaster? Do you wear a seat belt? Run all commands as root?
The history of computing says that everything is a porous security boundary just waiting for a motivated individual to attack it. Defense in layers, especially in the wild west of AI where everybody is playing move fast and break things at breakneck speed.
I understand some of the motivation for this post but in my view it's exaggerated because Claude code does a good job asking for permissions and in the (reasonable)worst case you can probably get an old state from github. If for some reason Claude (or someone) wants to take over your whole system I'm not sure if Dockers would stop it
I started working with VSCode in container ~3 years ago [1], at first I just felt I'd like to have control over the environment but later it made more sense when I switched to codeium windsurf [2]- I thought it's safer not to let agent loose in my filesystem
I put in a full day trying to get Claude Code and VSCode to work inside Docker. I wasn't able to get the window to properly display in my wayland session.
I don't know if it's appropriate for your case, but devcontainer <https://devcontainer.community/awesome>, code-server, or Eclipse Theia are basically VSCode over http and will mimick the GitHub.dev experience. I'm guessing a lot of your heartburn was trying to get the graphical parts of Electron to operate inside docker but it's really not good at that
I did see the VSCode part, but it might interest you that the JetBrains products also have an rpc style development setup through Gateway, where the product runs effectively headless in the container and just sends the display commands to your local IDE to handle latency sensitive GUI operations
You need to share the wayland socket (how programs talk to the compositor) with the container in docker. I’ve seen a script that does that on GitHub but I can’t remember the name. Distrobox and Toolbx also does this, but they also share your home directory with the container because that’s where all your config are.
I can highly recommend bubblewrap for sandboxing -- configuration is just CLI arguments, one per a mounted path; I use a little script that mounts CWD, claude stuff, and XDG data directories in RW, and the rest in RO.
To me it seems like the only reason to fiddle with docker is if your dev env is already in a container
I use DevPod [1] to run Claude Code within Docker-based dev container environments, and it's been fairly seamless. It lets me run environments locally and focus Claude's context on just the repo I want it to work on.
Bluefin distro is container/isolation oriented (sort of required given the immutable nature and not wanting to do too much layering) and includes a lot of options to implement a dev environment, including devpod and distrobox/toolbox.
Another idea is to develop in a KASM rootless docker workspace, where distrobox/toolbx can also be at your disposal. KASM workspaces are accessible thru the web, so you can work from anywhere.
The immutable nature of ideas like Bluefin and KASM encourage the user to use brew for package management, which would stay persistent under /home/linuxbrew
Another advantage of running claude inside a container is that you can use "--dangerously-skip-permissions" and let it do a thing in full autonomy for 2 hours. With a good prompt, abundant feedback like centralized logging and tests and a clear Claude.md, it can get pretty far on its own.
What drives this degree of user-experience complexity? Is there a targeted workflow that VsCode and Claude did not include in their GUI, that motivates this solution. Is it a tinkerer mindset, where setting this up is a fun activity for the author and readers?
This is my approach but it wasn’t all smooth especially with MCPs. And Docker wants you to install their Desktop software which is crap. I am not touching that thing. Yes, devcontainer extension is a thing but it’s a simple tool. We need something robust OOTB.
Mount them to the local filesystem, maybe even as read-only so Claude cannot modify them.
It actually makes bootstrapping the dev env so much easier too, basically just take a template .devcontainers and you're golden. After that just tell Claude what files you want to mount to the filesystem and restart
i don’t get the worry. i run these models all day without any sandbox and even leave them running while i walk away. i haven’t had a rm -rf kind of situation ever or even a hint of model going towards it. even gemini 2.5 at it’s lowest doesn’t do that.
Just because unintended things aren't happening right now, doesn't mean they won't happen. We are in the honeymoon phase of this technology where mass exploitation isn't yet being attempted.
However, if you are familiar with Pliny the Liberator's work, essentially all modern models are easily jailbroken, such that the original prompt can be overridden. All it will take for your agent is to download a malicious payload, perhaps disguised as a relevant library or documentation for the task at hand, and it can be running whatever the attacker tells it.
An 'rm -rf /' would be a pretty mild outcome. The more likely one would be the attacker silently installs malware on your machine.
I mistakenly started Claude in my root dev folder in dangerously skip mode the other day. Asked it to fix my playwright tests. All hell broke loose and I couldn't stop it. Lol. Eventually got it shut down and all was fine. Scared me though. There's a hook you can easily implement for RM -rf, see indydevdan on YouTube, he has a nice repo of it.
I would like to run everything in container and fire up two or three CC at once. It might be best actually. Right now I'm doing local dev with many services in one container and Claude sometimes doesn't use the container properly. I probably waste a lot of cycles with CC relearning my setup often as it struggles through it.
same, it actually annoys me when some calls need sudo access and I need to handle it manually, I have no idea what people are doing with those tools to fear destructive actions on the system
Fear and resentments. I've seen this not just online but also in conversation with friends. They encouter blogs and Youtube vids about great success with AI agents. Mainstream press articles about how AI is set to eliminate all jobs in 3 years, and mass layoffs in tech because 'AI' (even though we have had the same cyclic boom/bust in tech every 7 years before 'AI', not saying this cycle isn't different though).
They even tried some ChatGTP or Copilot coding themselves, throwing in some vague 'do this' prompt and got bad results. So they decided either "it's all bs hype for losers' or "man, I'm clearly not cut out for this new AI world, hope I can hold out till I can retire"
So every time they see an AI post, they feel the need to wave a '100% pure non AI coder, who's with me?' flag.
Yeah, I don't think you deserve the down votes for expressing an opinion (except maybe that it doesn't add a lot to the conversation), but speaking from personal experience, the whole AI coding thing has been pretty amazing for me.
I'm technical, I can read code (even write some of it), but actually completing things or getting started on serious projects is really hard: partly because I am not a professional develop (and so everything is hard) and partly because of time constraints. AI agents address both of these concerns very well and I've been able to produce things I wouldn't have thought possible. Are they ready for Prod? Maybe not, but at least they WORK and that's soo much more than anything I've managed in the past.
I swear to you that if you program the way I do you (without any AI tools or even developer tools made after 2005) you will learn 100x more and achieve 100x more
Except you glossed over the part where the person said they lacked time.
There is programming the art and programming to solve a problem. The art is great - I’ve written software for decades. Life dictates I no longer have as much time, but I still want to solve problems. AI helps me do that in the limited that I have.
The fatal flaw in your theory is that Docker, Claude, and whoever else will add an infinite amount of time spent debugging once you have a sufficiently complex problem, which is usually within 1-7 days of using it.
Please explain to me how I'm wrong. in the event you don't want to hear how I'm right, I am forstalling that while waiting for your reply
This thread is full of examples of people using the tools successfully. I'm also not sure where you're getting the 1-7 days from.
And someone who has the time may learn more by doing things without AI, but that's not what's being discussed. The person you originally responded to and I said we don't have the time.
It's a bit like arguing that AI audiobooks shouldn't exist because they are worse than human read. That may or may not be true, but that's not the discussion. AI may be the only reason an audiobook exist at all for certain books. AI coding is often the same way where someone can build a lot more than they could prior, even if it's imperfect, it may still solve their problem.
I agree with you that I would learn more - without any doubt. But there is no way that the second part (the achievement) stacks up. Over the weekend, I deployed a .Net Core API application, developed on Ubuntu and deployed on Windows Server Core, that includes a detailed test suite, full swagger documentation and which works exactly as I had hoped. Up to that point, I had never touched .Net Core, nor Windows Server Core. There is literally no way I would have been able to do that without the assistance of AI agents. Maybe you could, but I couldn't.
This is the fatal mistake you are making. What does the application in question do?
Do not mistake the mix of technologies you have used with the problem the code solves. One is a meaningful thing to focus on, and the other is something that career underachievers have invented to keep their salaries.
If you are a VSCode hater (like me), you can achieve the same thing using the open-source, Microsoft-built, official devcontainer CLI [0], which is exactly what VSCode uses under the hood. You don't get automatic port forwarding with this tool, which includes SSH Agent forwarding, so you'll likely want to pair this with installing SSH into your devcontainers, or using some other tool that does the job. I built a tool [1] to do the port forwarding part, which I'd like to be more magic like VSCode's but haven't put the effort into doing, since it works for my purposes.
Using these, I can SSH into my devserver, use the devcontainer CLI to open a tmux session with iTerm2 Tmux integration, and it feels just like a local window. Combine that with Neovide and remote Neovim over the same channel, and I have a native-feeling dev environment running on a remote devcontainer in a nearby data center.
[0] https://github.com/devcontainers/cli
[1] https://gitlab.com/CGamesPlay/qtm
Not loving that the devcontainers CLI is node. The appeal of devcontainers is to limit the amount of code I have to trust on my host. Give me a small static binary I can deploy to new systems.
I ssh into my server, then use the devcontainer cli tool to open a shell in the dev container. Found that easier than installing ssh in every container, especially when there's lots of them. I made a simple command line tool for helping the process, so I don't have to type out any long commands. If I run it it will list the running dev containers and I give them each a number to I don't have to use the full ID.
Yeah, this works. Two drawbacks: you have to rebuild the container when the port forwarding configuration changes, and the port forwards have to be unique on the entire machine.
I like the CLI for commands, but would rather work with an agent chat interface for vibing, which unfortunately I’m becoming increasingly reliant on, even though it’s gotten me into trouble with larger projects.
I have many environments locally, some dependent on others, and some require local databases. I use containers in production, but not always locally.
It’s almost a hellscape situation for trying to setup a fully isolated dev environment. I might be able have have some future generation of Claude convert my existing dev setup to something isolated with k8s. But, I don’t have the power to run such an environment locally, with all of the apps and services. I’d need to set my dev environments up in the cloud, but I don’t have enough money to pay for that, and it wouldn’t even work in some circumstances, because some things must run locally.
So, instead I run Cursor AI with various MCP servers, I approve tool calls, I don’t have things isolated, and I know this will eventually bite me.
> which unfortunately I’m becoming increasingly reliant on, even though it’s gotten me into trouble with larger projects
You could… not do this. You know it causes you problems, you consider it unfortunate that you’re becoming more reliant on it, and yet for some reason you’re choosing to continue anyway. Why are you doing that?
In a similar situation here: it's "unfortunate" in one sense but "extremely fortunate" in another sense.
It's unfortunate if you consider your primary craft to be writing code, as those skills can atrophy. It's obviously important to review and understand all material produced by these tools.
It's fortunate if your primary organizational role is "fix problems, communicate technically, and make new things" because of the sheer power of these tools when applied properly.
I've sunk many hours into attempting to "claude-ify" a complex set of applications and services and (IMO) that's a very useful activity.
Why? So that I can use ai-tooling more effectively? Yes. Why also? So that services are more decoupled, more testable, more aligned with good development principles.
It's difficult to inject these workflows in a useful way across multiple levels (code/devops/org) but when it works it's worth it.
My key takeaway was something along the lines of: "if an agent can't understand and work with your codebase, you've got an onboarding problem" (we do).
Disclaimer: mostly standard web tech across Java/Scala/React - aggressively complex k8s layer
You might as well ask why people addicted to gambling don’t just stop.
devpod is quite nice for this
https://devpod.sh/
Just tried it... and it says it's client-only, but when I try to run the SSH provider using my Rapberry Pi it says it needs `docker` to be installed on it :/. So it's not just client-only, is it?
It's client-only because if you install it on a machine with Docker, you don't need a server to use devcontainers.
Docker is required to use devcontainers, so there's really no getting around the Docker requirement
Isn't that what client-only means?
No, client-only means Devpod is contrasting itself to Coder, which requires a central server to orchestrate the containers it creates. Devpod does not. It does still have dependencies, though, among them Docker.
Like, oh you need to install this extremely heavy tool on your server... but yeah it's still client-only, trust me?
No, client-only as in this is the client only. You need a corresponding server. Are you thinking of "runs locally"?
Mate, if you need to install something on a server how the hell can it be called client-only??? eMacs is an example of client-only perhaps, with TRAMP as it works purely on top of SSH , though you could argue it still isn’t client-only as it requires the SSH server to be enabled… so I thought the SSH connector for this would also work the same way but it also requires a full freaking Docker install not just SSH. This is totally misleading.
Oh the gitlab link is really really really cool. I can't express in words how happy I am since I know about iroh and I always wanted to create such a tool myself too but never could figure out iroh.
Now the way I did port forwarding was really wild and really complicated and I think that it might demand a blog post if I am being honest, but yes this project is absolutely godsend. Thanks!!
> If you are a VSCode hater (like me)
What are some reasons to hate VSCode?
It's proprietary software masquerading as open-source, for one, and intentionally fosters dependence on its vendor for the plugins that support the most popular languages as well as key features (remote editing, WSL support, integration with the mainline plugin repos), for one.
But Cursor, a fork not developed by Microsoft, ships with remote editing, wsl support, etc
It's their reimplementation.
That’s not what that word means
They do indeed use a different implementation of these features than the one that ships with VSCode, though they're not the original authors of it. They used to host their version separately here: https://github.com/anysphere/open-remote-ssh
But I guess they've just rolled it into the Cursor repo? Idk, because Cursor itself is proprietary.
There's a similar extension for WSL integration, though I'm not 100% sure if Cursor's implementation of this feature is based on it: https://github.com/jeanp413/open-remote-wsl
You can sideload proprietary VSCode extensions to add them to VSCodium or some downstream fork, but you may also have to patch the extensions themselves because some of them refuse to run if they aren't being used with Microsoft's VSCode: https://github.com/OliverKeefe/vscode-extensions-in-vscodium
Microsoft injects proprietary spyware into the official builds that you can't see from the source code. You can get unofficial builds of just the source code from the VsCodium project, but Microsoft blocks your ability to install a lot of extensions if you don't use the official build.
It’s clumsy as hell. It’s neither a good editor, which focus on being fast when doing things with text, nor a good IDE, which is more about the tooling and getting a project done. It’s trying to be both at the same time. And the result is bad at both.
I don’t know, others might be better out of the box. But once you start adding features they all become a mess. I rather just deal with a known mess and not waste time with tools.
I have nothing against VSCode but I don’t use it.
For small files and quick edits i use neovim with no customisation at all. Back when I first started using vim I had a pretty long config and some plugins installed and so on. But now that I no longer try to use it as an IDE, neovim is perfect for small edits.
For development work I use JetBrains suite of tools. Very minimal amount of customisation there also, only a couple of extra plugins installed by me and a couple of changes in the settings. It works well enough that I don’t have any desire to spend time customising it a whole bunch.
Don’t get me wrong, I loved neovim and other tools but tbh I just nano for simple edits now. AI tab complete is just too good. Basically it’s barebones or VSCode (or its variants). I could probably use vanilla neovim instead of nano but I've not used vim all that much recently to be in that "flow".
This is like saying: I can write Swift code using Vim instead of Xcode. Yeah, you can, but that doesn’t mean you should.
Note: I use nano/vim daily to read and edit files on my servers. Just not to build complex apps.
Your point is good, I don't mean to undermine it, but because I work in this area I feel compelled to point out that depending on the type of Swift you're writing using NeoVim to write Swift is actually quite pleasant: https://www.swift.org/documentation/articles/zero-to-swift-n...
I'm hoping one day to move from a GUI to a VUI (Voice User Interface), though I still have a soft spot for TUIs. Thanks for the link :)
I'd waste time with tools. I have my vim config, my git config, my emacs config, my bash config,... that I copy over to any development environment I have. And every now and then, I go into the manual or other people config to check if there's any new trick I can apply.
Comfort is something invaluable when doing a task. I can accept discomfort when I'm in an unfamiliar environment, but not when it's going to be something I will be in for days.
Perhaps the difference is that VSCode and plugins change too often and requires too much maintenance?
Ultimately, they're about as trite as reasons to hate Vim or Emacs. It boils down to preference, and I'm outlining a path to using another editor and getting access to some of the same features.
Each time I try to use it, I find myself distracted by the impulse to fix the bugs that I notice. There's always this moment when I find a comment from myself, several years earlier, in an issue that's still open because it's unclear whether it's a VSCode problem or a plugin problem, and it's like: "oh, here I am again, doing this instead of working".
That's not to say it's especially buggy, it's just so large, there are a lot of places for bugs to hide.
I get distracted by the impulse to contribute to other editors also, but those are contributions that I'm happy about. They don't always result in bugs fixed or features added, sometimes somebody says "no" but at least it resolves in some way. With VSCode it feels like I'm going in circles.
I personally feel some inner sense of dissonance when I use software written in Electron when alternatives are available. Software targeted to the technical audience has no valid reason not to aim higher in terms of performance and fundamentals.
I will give you one reason why I don't like it much:
- The push to use Github Copilot and no option to disable that crap.
I wish there was a fork of VSCode with all the telemetry and all of the proprietary crap disabled with 100% extension support...
I do not hate it when others are using it (none of my business). But I do not like to use bloatware. Text editor that eats RAM (even excluding LSP) like 3d editing software.
Microsoft ?
Through the use of amazing PR they have made people forget about their Embrace, extend, and extinguish strategy with open source.
GP went on to recommend Microsoft devcontainers, so I presume that's not the reason.
VSCodium - Free/Libre Open Source Software Binaries of VS Code
https://vscodium.com/
Just so you know, the devcontainer VSCode extensions are considered proprietary by Microsoft and therefore not allowed to be installed in VSCodium. There are some open-source container plugins, but they are all a bit janky for the container use-case in my experience.
Thanks for this. I didn’t know they shipped something like this.
Thanks
I have actually been working on something similar but instead of Docker, I am using Apple’s builtin container[1] support to run shell commands and code with real OS-level isolation. It’s fast (spawns in milliseconds) and integrates nicely with things like Claude Code and the Gemini CLI. I open sourced it as CodeRunner[2]. Would love to hear what people think or chat about how it compares.
1. Apple container: https://github.com/apple/container
2. CodeRunner: https://github.com/BandarLabs/coderunner
Caveat: You need M1/M2/M3/M4 mac for this to work. MacOS 26 is optional but recommended.
TIL you can just run the container command with an existing docker file
I have enjoyed running Claude Code in a container. The biggest advantage for me isn't security though, it's how easy it becomes to spin off agents to work in the background.
I use a simple script that copies my working directory into a container, prompts Claude Code, and then either saves a planning document locally, or opens a pull request in GitHub for me to review.
I quite like this because it makes starting agents when I have ideas really frictionless, and then I can easily discard the results and try again if it went off the rails. The Max plan makes me not worry about the cost of this whole process as well.
I also set up containers initially out of caution, but honestly, I've never seen Claude Code do anything that git couldn't easily reverse, so I'm not that worried about that any more.
And thus, the machines managed to build trust in their harmlessness in men. Which, eventually, would turn out advantageous..
Yeah the point of security is missed. Given how easy it is, I doubt I'm ever going to run Claude code or any other CLI that uses remote AI servers without a devcontainer again
> I've never seen Claude Code do anything that git couldn't easily reverse, so I'm not that worried about that any more.
I had one agent that installed something with pip and used it to rewrite my git history. It used the wrong flag and what was left was one file. I am glad I caught it before auto-commit and push triggered. It was one of my hobby projects, so not a huge deal. I can't imagine the fallout of this happening in a work environment.
Have you tried setting up multiple git worktrees as well? I wonder if that doesn’t solve most of the issues.
I have, but I just found that a bit tedious to manage manually. Maybe another version of this script would just be to automatically create a worktree, run a prompt in that, and then have the script manage the results, and clean up the worktree at the end.
One thing that I’ve found works well is to permanently have a few numbered branches checked out as worktrees, keep them up to date with main, and then when you want to do a task in parallel you just pick an existing branch rather than creating and deleting worktrees every time.
This lets you persist local databases, IDE project settings and so on for each worktree too.
https://github.com/stravu/crystal
This is still alpha, but I have been using it to handle the multi-agent worktree paradigm.
It does, yes, and it's what Claude Code recommends officially.
Even locally, Claude in multiple terminal tabs/windows can do different tasks in parallel
They can, but I find it harder to review if you are trying to get Claude Code to do anything non-trivial. I've seen some people make multiple git worktrees so they can have two instances of Claude Code working independently, but I found that to be tedious to manage.
Isn't there tooling for helping to manage this
Any interest in sharing the script?
Not the OP, but I have this fish alias. I just run `claude-docker` in my project root (where I would otherwise run `claude`). There's nothing to it. You can tweak it:
I'm afraid it is not that shareable as it contains a lot of dependency management and setup that is specific to my setup and projects. But it's not too complicated. You could probably re-create your own version fairly quickly.
It's just a Dockerfile that installs dependencies, an entrypoint that invokes claude, and some wrapper scripts that handle creating the container, passing through the prompt, and managing the results (e.g., output markdown to a directory, or open a PR). For input/output from the container I mount directories within the Docker container to local directories under ~/.call_claude/. I create a directory under there, clone the source code to it, output the prompt to a text file, and then grab results from that directory after Claude Code has finished running.
Just have Claude make a version of your script that's shareable for others, no?
It's just not that simple. If your scripts contain 20 things hard-coded to your setup and projects, it is not quick nor easy to remove all of those and make something configurable and easily adaptable to other people's setups and projects.
Adapting the scripts to your specific setup is pretty much the entirety of the work in setting this up.
Yes please, more containers-first thinking for all things AI. The fact that this has been largely absent from most discussions and demos regarding agents and vibe-coding seems like a big red flag to me. I mean if we're going to YOLO the code, perhaps we should at least pay some attention to the dev / test harness?
It's worth thinking about reproducibility even if you're not concerned about security. We would not have to argue so much about the effectiveness of something like vibe-coding if everyone could simply share process/tooling without tons of custom setup. And even if you're not worried about filesystem sandboxing and leaking secrets like ssh keys, even if versioning with git is enough of an undo button for you.. who wants to clutter their whole system with a ton of new language stacks and/or weird dependencies for every small experiment? Most experiments with the new hotness are going to fail, so we should expect to do a lot..
Containers bring their own set of problems, there are some examples brought up in this thread, mainly around communication with the host OS.
I‘d argue the reproducible parts of vibe coding (agentic engineering) setups are just text files. Many people use a mix of web apps (AI studio), Mac apps (Wispr Flow), and other UI tools (repo prompt) in their workflow which can’t be put in a container anyway - well, reasonably at least.
If you want security, containers won’t get you that far. You’ll need to use a VM.
But if you give Claude Code access to your GitHub repo, what else is there worth protecting, that’s not protected by keychain & sudo?
All development is text files, that is missing the point. The development environment is a system, and a pretty complicated one too. It matters where the files are, what's in them, and how they interact. Things change together instead of staying isolated, you add more pieces over time, and even more things need to change together. Anyone who likes text-files more than click-to-configure UIs for tools, will probably like containers more than text files for systems, and for all the same reasons.
Your choices to reproduce complex systems are basically to 1) deny that complexity exists and accept any corresponding limitations in your working environment, 2) follow some error-prone multistep processes to reproduce existing system setup manually, 3) commit to centralizing all development on external cloud platforms, or 4) do something else to bundle together a setup that's repeatable.
I'm strongly in favor of (4) here, and while I'd probably disagree that it requires VMs instead of docker, the goal of repeatable processes is so important that hey, whatever works. It sounds like you're in camp 1 or 2?
> The development environment is a system, and a pretty complicated one too.
It's also partly personal. I also like to use isolated environments, mostly because I tinker with different technologies, and over times, it becomes a mess on my machine.
But I still like my git config, my editor and other tooling that I have to assist in my tasks. And it's not worth it to replicate it over to other people's setup. And neither I want their things.
> But I still like my git config, my editor and other tooling that I have to assist in my tasks. And it's not worth it to replicate it over to other people's setup.
This isn't about stopping you from customizing your development environment. It's about making your custom development environment automatically useful on your desktop or your laptop, or committing a project-specific starter environment to version control so that it doesn't take everyone on your team days to onboard.
> And neither I want their things.
Why not? This isn't about something like git aliases or editor keyboard shortcuts. It's about tooling pipelines. If someone's claiming enhanced productivity I don't want to watch their 1 hour youtube video, pause it, and painstakingly recreate their setup to check if the claims are true.
There's a reason projects grow Tiltfiles, compose files, and similar. In the limit, agents are basically just a collection of microservices that just happen to operate on code. How do you expect to do that sort of thing without tools?
And you don't need to stop there. Claude Code also has great open alternatives you can run in any environment you like, with any model, and under economic arrangements you define. See https://oss-ai-swe.org/ for a few. My current favs are OpenHands and Codex CLI.
Aren’t those _way_ more expensive per token?
Yes and I really hope more models consider pricing like Claude Max. There's also the option of choosing local models but I haven't tried that yet. Claude is just too good to spend time elsewhere (at the moment).
TANSTAAFL
https://everything.intellectronica.net/p/all-you-can-eat-eco...
How so? Claude is the most expensive. With open tools you can choose whatever model delivers the best results for the budget you have.
you can use your claude code subscription with opencode (not sure how "legal" that is); github copilot straight up provides an api (very "legal")
Containers do seem to work really well for agents. It lets them run in parallel! Giving them sudo also lets them do interesting things you wouldn’t want them doing on your desktop, eg. installing tools or debugging with tcpdump. (I am working on agent-in-a-container: sketch.dev)
Letting Claude Code loose on my home folder feels bad yes. Would using a separate user account solve it? I.e. keep git repos in for example /home/claude and run claude as that user? Or am I missing something?
I do pretty much this.
- /home/me -- 660 default permissions
- /home/me/project -- owner "claude", add "me" to "claude" group
- symlink /home/claude/project to /home/me/project
- Run claude in different user session
- Use CTRL+ALT+F2 to immediately switch between my main and claude sessions. Claude continues working in the background.
This is exactly why I’m using chatGPT codex over Claude Code (even though I suspect CC might be better). Each chat runs in its own cloud based container. There’s no risk of exposing secrets/deleting files/breaking environment. I can run 3 in parallel while I continue to code locally. It works in a branch and integrates via PR. It’s a great DevEx.
Someone please build a nice macOS app using hypervisor framework that lets me do this locally :)
Honest question: why do people prefer developing code inside a docker? I get the benefits of docker as a deployment unit, but wouldn’t configuring a dev container and using it a hassle nonetheless, compared to not doing them at all?
Personally I don’t trust running npm/pip/gradle/etc in my own machine. I feel more secure running them in containers or VMs. Some IDEs like jetbrains’ support remote development (it’s not perfect but better than risking my own machine)
Do you feel safe driving in traffic / breathing polluted air / having restaurants prepare your food?
Clown emoji.
If you commit the .devcontainer.json and associated files your whole team gets a consistent, local-feeling, environment for free.
Containerisation solved the "Works on my machine" gap between local and prod. This is just solving for the same gap between Dev 1 and Dev 2.
That's a different thing. You're talking about running the software under development in a container. I think the commenter was asking about running the dev tooling in a container, which I find a massive pain.
No, that’s what they’re talking about - running the tooling in a container. Dev tool chains are a pain in the ass to keep consistent in even small teams. Devcontainers solve that.
Devcontainers don't "feel" local to me, since my CLI tools are not available in them. Think ripgrep or fd, for example.
You could include those as part of the tooling. I have been experimenting with including mise as part of the image and then layer on the extra tools within there. Put all of those steps into the build so it is automatic.
Separation of dependencies and ability to easily interconnect containers.
If you’re like me and want to maintain a clean dev machine. I don’t want my machine to get in my way when I need to get work done. That means I don’t want some X messing things up somewhere that I didn’t know about. Older I get my memory isn’t as sharp and things like this matter.
If the host OS is not Linux (usually Windows or macos) then "running in a container" is cover for "running on the right OS".
I thought it was more commonplace to develop inside containers or vms. I don’t trust running npm or pip or gradle or go install in my own machine.
I use VMs myself. I use Proxmox and have it setup so that I can spin up a fresh VM in around 10-20 seconds when needed. I also like that I can take snapshots.
I really want to limit the amount of software I am installing on my main system due to possible security issues.
Do you have any example scripts for your setup?
I have recently moved to a VM only development workflow, but still feeling some growing pains as I figure out a good setup. I run my IDE in the guest, which is kind of crummy - responsiveness is worse + the VMs are a lot heavier than a headless code setup would be. I have thought about using a Flatpak IDE with filesystem access disabled as a potential middle ground solution so that the guests could be minimal images connected by SSH.
I have taken cloud images and set them up with 'cloud-init'. So I can bring them up very quickly.
If you are using Claude Code on macOS running it in a devcontainer has a few upsides like fewer cli tool call failures, meaning less waiting time and less context window spam.
But there are some things you lose as well @ ergonomics:
- "done" notifications are not working by default (might be possible trigger and send to host using hooks feature)
- more difficult to give playwright a session (need to get cookies into playwright in the container)
- can't easily paste screenshots into the devcontainer (Claude Code has to take a screenshot or you have to put it into a mounted dir)
and of course the usual advantages of using containers still apply (dependency management, reproducibility and so on).
I still wonder though if nix would allow me to remove the container overhead (they do noticeably tax the macbook battery).
I never have tool calls failing on macOS. What tool calls are failing for you?
Claude Code using grep is often failing (also in devcontainer but less often). Many command line tools on macOS work subtly different than the ones on linux.
can't you easily circumvent this by telling it to use the Mac versions in the prompt.md
This setup is also compatible with GitHub CodeSpaces, which I think is an exceptionally nice development experience if you are on something like a Chromebook, and gives you isolation from your local network. Also lets you share your app running in dev-mode with anyone for feedback, without having to poke holes in your network or whatever. Also also, you can use the Cline extension in this environment if you still want the cursor-style copilotey experience.
From what I can gather, this is part of the idea/appeal behind something like https://phoenix.new/. The AI Agents have total control over the environment and can use all of the associated system tools as a result, without putting your own machine at risk. Once things are kinda working, you can pull it out and work locally.
2-ish questions:
Is this level of fear typical or reasonable? If so, why doesn’t Anthropic / AI code gen providers offer this type of service? Hard to believe Anthropic is not secure in some sense — like what if Claude Code is already inside some container-like thing?
Is it actually true that Claude cannot bust out of the container?
> Is this level of fear typical or reasonable?
Just a month ago, an AI coding agent deleted all the files on someone's computer and there was a little discussion of it here on HN. Support's response was basically "yeah, this happens sometimes".
forum post: https://forum.cursor.com/t/cursor-yolo-deleted-everything-in...
HN thread (flagged, probably because it was a link to some crappy website that restates things from social media with no substantive content of its own): https://news.ycombinator.com/item?id=44262383
Idk how Claude Code works in particular, though.
It's worth noting that the default settings of Cursor do prevent this by asking you to confirm every command that is run. And when you get tired of that 5 minutes in and switch to auto-approving there is still protection against files outside the work directory being deleted. The story above is about someone who disabled all the safeguards because they were inconvenient, then bad things happened
It is a good example of "bad things can happen", but when talking about whether we need additional safeguards the lessons are less clear. And while I'm not as familiar with the safeguards of Claude Code I'm assured it also has some by default
re Anthropic: there is a whole docs section on devcontainers
https://docs.anthropic.com/en/docs/claude-code/devcontainer
and an example repo
https://github.com/anthropics/claude-code/tree/main/.devcont...
--cap-add=NET_ADMIN
--cap-add=NET_RAW
https://github.com/anthropics/claude-code/blob/main/.devcont...
If the cointainer route (with those types of privileges) is being suggested from a security point of view, then might as well run these processes in a VM and call it a day :/
I haven't found that to be the case. I have used cc within an container and on the host machine and it has been fine. Any command that could cause changes to your system you MUST approve when using it in agent mode.
You either have the option of approving each command manually, or you can let it run commands autonomously. If you let it run any commands then you have the risk of it doing something stupid (mainly deleting files).
You also have MCP tools running on your machine, which might have security issues.
I have personally never seen claude (or actually any AI agent) do anything that could not be fixed with git. I run 24/7 in full permissions bypass mode and hardly think about it.
Correlation does not equal causation as the old adage goes. Just because if you havent seen the pattern, doesn't mean it can't.
It is like insurance, 99.95% of the time you don't need it. But when you do, you wish you had it.
> Is this level of fear typical or reasonable?
Anyone with more than one toolbox knows that fear isn't required. Containers are about more than security, including stuff like organization and portability.
> If so, why doesn’t Anthropic / AI code gen providers offer this type of service?
Well perhaps I'm too much the cynic, but I'm sure you can imagine why a lack of portability and reproducibility are things that are pretty good for vendors. A lack of transparency also puts the zealots for "100x!", and vendors, and many other people in a natural conspiracy together, and while it benefits them to drum up FOMO it makes everyone else burn time/money trying to figure out how much of the hype is real. People who are new to the industry get leverage by claiming all existing knowledge does not matter, workers who are experienced but looking to pivot into a new specialization in a tough job market benefit from making unverifiable claims, vendors make a quick buck while businesses buy-to-try and forget to cancel the contract, etc etc.
> Is it actually true that Claude cannot bust out of the container?
Escaping containers is something a lot of people in operations and security have spent a lot of time thinking about long before agents and AI. Container escape is possible and deadly serious, but not in this domain really, I mean all your banks and utility providers are probably using Kubernetes so compared to that who cares about maybe leaking source/destroying data on local dev machines or platforms trying to facilitate low-code apps? AI does change things slightly because people will run Ollama/MCP/IDEs on the host, and that's arguably some new surface area to worry about. Sharing sockets and files for inter-agent comms is going to be routine even if everyone says it's bad practice. But of course you could containerize those things too, add a queue, containerize unit-tests, etc
>Is this level of fear typical or reasonable?
Of course. Also with regular customer projects. Even without AI--but of course having an idiot be able to execute commands on your PC makes the risk higher.
> If so, why doesn’t Anthropic / AI code gen providers offer this type of service?
Why? Separate the concerns. Isolation is a concern depending on my own risk appetite. I do not want stuff to decide on my behalf what's inside the container and what's outside. That said, they do have devcontainer support (like the article says).
>Hard to believe Anthropic is not secure in some sense — like what if Claude Code is already inside some container-like thing?
It's a node program. It does ask you about every command it's gonna execute before it does it, though.
>Is it actually true that Claude cannot bust out of the container?
There are (sporadic) container escape exploits--but it's much harder than not having a container.
You can also use a qemu vm. Good luck escaping that.
Or an extra user account--I'm thinking of doing that next.
I don't get these posts. I'm using claude --dangerously-skip-permissions all day and haven't had a single issue. In my experience it doesn't just randomly start erasing your hard drives. Also having a proper CLAUDE.md probably helps.
> it doesn't just randomly start erasing your hard drives.
if it did, it would be too late, wouldn't it?
> I'm using claude --dangerously-skip-permissions all day and haven't had a single issue.
Yet. You haven't had a single issue yet.
Correct. The number of times I’ve had Claude ask permission to rm -rf paths with .. elements is more than zero.
I have in fact had Claude (running on Cursor) do the equivalent of rm -rf /. It happens.
You ignore safety features unless you have already experienced disaster? Do you wear a seat belt? Run all commands as root?
The history of computing says that everything is a porous security boundary just waiting for a motivated individual to attack it. Defense in layers, especially in the wild west of AI where everybody is playing move fast and break things at breakneck speed.
I understand some of the motivation for this post but in my view it's exaggerated because Claude code does a good job asking for permissions and in the (reasonable)worst case you can probably get an old state from github. If for some reason Claude (or someone) wants to take over your whole system I'm not sure if Dockers would stop it
Some people never create backups and they never had a problem either.
I started working with VSCode in container ~3 years ago [1], at first I just felt I'd like to have control over the environment but later it made more sense when I switched to codeium windsurf [2]- I thought it's safer not to let agent loose in my filesystem
[1] https://github.com/grzegorzk/vscode_in_podman
[2] https://github.com/grzegorzk/codeium_windsurf_in_podman
I put in a full day trying to get Claude Code and VSCode to work inside Docker. I wasn't able to get the window to properly display in my wayland session.
I don't know if it's appropriate for your case, but devcontainer <https://devcontainer.community/awesome>, code-server, or Eclipse Theia are basically VSCode over http and will mimick the GitHub.dev experience. I'm guessing a lot of your heartburn was trying to get the graphical parts of Electron to operate inside docker but it's really not good at that
I did see the VSCode part, but it might interest you that the JetBrains products also have an rpc style development setup through Gateway, where the product runs effectively headless in the container and just sends the display commands to your local IDE to handle latency sensitive GUI operations
> but devcontainer <https://devcontainer.community/awesome>,
404
Here's the intended link target: https://github.com/devcontainer-community/awesome-devcontain...
You need to share the wayland socket (how programs talk to the compositor) with the container in docker. I’ve seen a script that does that on GitHub but I can’t remember the name. Distrobox and Toolbx also does this, but they also share your home directory with the container because that’s where all your config are.
Are you using devcontainers? The vscode UI doesn’t run in docker, but a remote server does.
I can highly recommend bubblewrap for sandboxing -- configuration is just CLI arguments, one per a mounted path; I use a little script that mounts CWD, claude stuff, and XDG data directories in RW, and the rest in RO.
To me it seems like the only reason to fiddle with docker is if your dev env is already in a container
I use DevPod [1] to run Claude Code within Docker-based dev container environments, and it's been fairly seamless. It lets me run environments locally and focus Claude's context on just the repo I want it to work on.
[1]: https://devpod.sh/
Bluefin distro is container/isolation oriented (sort of required given the immutable nature and not wanting to do too much layering) and includes a lot of options to implement a dev environment, including devpod and distrobox/toolbox.
https://docs.projectbluefin.io/bluefin-dx/
Another idea is to develop in a KASM rootless docker workspace, where distrobox/toolbx can also be at your disposal. KASM workspaces are accessible thru the web, so you can work from anywhere.
The immutable nature of ideas like Bluefin and KASM encourage the user to use brew for package management, which would stay persistent under /home/linuxbrew
https://hub.docker.com/r/linuxserver/kasm
https://kasmweb.com/docs/latest/how_to/docker_in_kasm.html
https://gist.github.com/jgbrwn/28645fcf4ac5a4176f715a6f9b170...
Interesting. Works with claude max or API only?
I use Claude Max with it. Haven't personally tried the API yet.
Another advantage of running claude inside a container is that you can use "--dangerously-skip-permissions" and let it do a thing in full autonomy for 2 hours. With a good prompt, abundant feedback like centralized logging and tests and a clear Claude.md, it can get pretty far on its own.
What drives this degree of user-experience complexity? Is there a targeted workflow that VsCode and Claude did not include in their GUI, that motivates this solution. Is it a tinkerer mindset, where setting this up is a fun activity for the author and readers?
This is my approach but it wasn’t all smooth especially with MCPs. And Docker wants you to install their Desktop software which is crap. I am not touching that thing. Yes, devcontainer extension is a thing but it’s a simple tool. We need something robust OOTB.
Try colima or rancher desktop
This is the idea I had in mind while doing https://github.com/textcortex/claude-code-sandbox
Wait, Claude Code isn't even open source? What year is it?! And people are letting this thing run commands directly on their workstations? Crazy...
How can I preserve the local configuration files I have with this container? (the custom slash commands, config.json etc)
Mount them to the local filesystem, maybe even as read-only so Claude cannot modify them.
It actually makes bootstrapping the dev env so much easier too, basically just take a template .devcontainers and you're golden. After that just tell Claude what files you want to mount to the filesystem and restart
I'm ashamed I didn't think of it. Great article, thank you!
Also can we please differentiate vibe coding and prompt based development.
There is a difference between prompting "write me a website" and just using the LLM for typing faster than you do.
i don’t get the worry. i run these models all day without any sandbox and even leave them running while i walk away. i haven’t had a rm -rf kind of situation ever or even a hint of model going towards it. even gemini 2.5 at it’s lowest doesn’t do that.
has anyone faced this?
Just because unintended things aren't happening right now, doesn't mean they won't happen. We are in the honeymoon phase of this technology where mass exploitation isn't yet being attempted.
However, if you are familiar with Pliny the Liberator's work, essentially all modern models are easily jailbroken, such that the original prompt can be overridden. All it will take for your agent is to download a malicious payload, perhaps disguised as a relevant library or documentation for the task at hand, and it can be running whatever the attacker tells it.
An 'rm -rf /' would be a pretty mild outcome. The more likely one would be the attacker silently installs malware on your machine.
I mistakenly started Claude in my root dev folder in dangerously skip mode the other day. Asked it to fix my playwright tests. All hell broke loose and I couldn't stop it. Lol. Eventually got it shut down and all was fine. Scared me though. There's a hook you can easily implement for RM -rf, see indydevdan on YouTube, he has a nice repo of it.
I would like to run everything in container and fire up two or three CC at once. It might be best actually. Right now I'm doing local dev with many services in one container and Claude sometimes doesn't use the container properly. I probably waste a lot of cycles with CC relearning my setup often as it struggles through it.
same, it actually annoys me when some calls need sudo access and I need to handle it manually, I have no idea what people are doing with those tools to fear destructive actions on the system
"I leave my door unlocked all day, sometimes when I'm not even home. I've never had anyone rob my house. Has anyone faced this?"
Thanks
[flagged]
Why is it that in every Claude Code related submission there's people proudly declaring out-of-context they are not using AI tools?
If there's a post about Zig, you don't see people saying how they don't use Zig.
Fear and resentments. I've seen this not just online but also in conversation with friends. They encouter blogs and Youtube vids about great success with AI agents. Mainstream press articles about how AI is set to eliminate all jobs in 3 years, and mass layoffs in tech because 'AI' (even though we have had the same cyclic boom/bust in tech every 7 years before 'AI', not saying this cycle isn't different though).
They even tried some ChatGTP or Copilot coding themselves, throwing in some vague 'do this' prompt and got bad results. So they decided either "it's all bs hype for losers' or "man, I'm clearly not cut out for this new AI world, hope I can hold out till I can retire"
So every time they see an AI post, they feel the need to wave a '100% pure non AI coder, who's with me?' flag.
I am proud that I still use pre-standardization C instead of Zig
Yeah, I don't think you deserve the down votes for expressing an opinion (except maybe that it doesn't add a lot to the conversation), but speaking from personal experience, the whole AI coding thing has been pretty amazing for me.
I'm technical, I can read code (even write some of it), but actually completing things or getting started on serious projects is really hard: partly because I am not a professional develop (and so everything is hard) and partly because of time constraints. AI agents address both of these concerns very well and I've been able to produce things I wouldn't have thought possible. Are they ready for Prod? Maybe not, but at least they WORK and that's soo much more than anything I've managed in the past.
I swear to you that if you program the way I do you (without any AI tools or even developer tools made after 2005) you will learn 100x more and achieve 100x more
Except you glossed over the part where the person said they lacked time.
There is programming the art and programming to solve a problem. The art is great - I’ve written software for decades. Life dictates I no longer have as much time, but I still want to solve problems. AI helps me do that in the limited that I have.
The fatal flaw in your theory is that Docker, Claude, and whoever else will add an infinite amount of time spent debugging once you have a sufficiently complex problem, which is usually within 1-7 days of using it.
Please explain to me how I'm wrong. in the event you don't want to hear how I'm right, I am forstalling that while waiting for your reply
This thread is full of examples of people using the tools successfully. I'm also not sure where you're getting the 1-7 days from.
And someone who has the time may learn more by doing things without AI, but that's not what's being discussed. The person you originally responded to and I said we don't have the time.
It's a bit like arguing that AI audiobooks shouldn't exist because they are worse than human read. That may or may not be true, but that's not the discussion. AI may be the only reason an audiobook exist at all for certain books. AI coding is often the same way where someone can build a lot more than they could prior, even if it's imperfect, it may still solve their problem.
I agree with you that I would learn more - without any doubt. But there is no way that the second part (the achievement) stacks up. Over the weekend, I deployed a .Net Core API application, developed on Ubuntu and deployed on Windows Server Core, that includes a detailed test suite, full swagger documentation and which works exactly as I had hoped. Up to that point, I had never touched .Net Core, nor Windows Server Core. There is literally no way I would have been able to do that without the assistance of AI agents. Maybe you could, but I couldn't.
This is the fatal mistake you are making. What does the application in question do?
Do not mistake the mix of technologies you have used with the problem the code solves. One is a meaningful thing to focus on, and the other is something that career underachievers have invented to keep their salaries.
[dead]
You’re missing out
[flagged]
This kind of behaviour is hilarious, either use the tools with their caveats, or actually make a stand by not using them.
60's movements would not be possible in current times.
"accept things how they are, or give up"
would not be possible, indeed
Standing up is not giving up, thanks for confirming my remark.