Show HN: OpenWork – An open-source alternative to Claude Cowork
github.comhi hn,
i built openwork, an open-source, local-first system inspired by claude cowork.
it’s a native desktop app that runs on top of opencode (opencode.ai). it’s basically an alternative gui for opencode, which (at least until now) has been more focused on technical folks.
the original seed for openwork was simple: i have a home server, and i wanted my wife and i to be able to run privileged workflows. things like controlling home assistant, or deploying custom web apps (e.g. our customs recipe app recipes.benjaminshafii.com), legal torrents, without living in a terminal.
our initial setup was running the opencode web server directly and sharing credentials to it. that worked, but i found the web ui unreliable and very unfriendly for non-technical users.
the goal with openwork is to bring the kind of workflows i’m used to running in the cli into a gui, while keeping a very deep extensibility mindset. ideally this grows into something closer to an obsidian-style ecosystem, but for agentic work.
some core principles i had in mind:
- open by design: no black boxes, no hosted lock-in. everything runs locally or on your own servers. (models don’t run locally yet, but both opencode and openwork are built with that future in mind.) - hyper extensible: skills are installable modules via a skill/package manager, using the native opencode plugin ecosystem. - non-technical by default: plans, progress, permissions, and artifacts are surfaced in the ui, not buried in logs.
you can already try it: - there’s an unsigned dmg - or you can clone the repo, install deps, and if you already have opencode running it should work right away
it’s very alpha, lots of rough edges. i’d love feedback on what feels the roughest or most confusing.
happy to answer questions.
Personally I’m skeptical and a bit dismissive of an insta- clone of commercial offerings. I ignore these things until they’re 3-6 months old and still iterating.
Announcements on their own are like whispers.
Completely understand. It's up to us to keep shippinh and making this actually good during that time!
respect the attitude!
What if the commercial offering was made in less than 2 weeks?
https://x.com/altryne/status/2010811222409756707
My take on the OP is that its commitment to an idea is what matters, not how quickly its created. I love seeing people insta-clone things but is this a side project that going to see updates for a few weeks or is this something that is going to be maintained actively for years to come.
It's a feature built upon a foundation. Cloning the feature without the foundation isn't the same thing.
The uncomfortable truth is that the marginal cost of producing software (if one doesn't care too much about quality or predictability) has now become near zero with AI; and therefore we will see many companies products being challenged in this way. Of course, Anthropic will have a leg up in terms of subsidized models provided along with a subscription, etc.
> the marginal cost of producing software (if one doesn't care too much about quality or predictability) has now become near zero with AI;
Hell of a caveat. I'd also add "complex functionality" to that list.
That's a good idea from the user perspective. I jumped on the aider bandwagon and am now dealing with its relative abandonment and deciding if I want to invest time in a fork.
As a movement though it might be something else. To me, Claude is trying to build a moat around tooling, but due to the nature of their own core product that moat is becoming easier and easier to bridge. What does that mean for them?
Looks great. I don't use opencode myself but like a lot Claude skills, agents, I have like 50 different ones for personal life, work, dating, whatever Is there integration with Claude skills?
Have you considered using or integrating with ACP? https://zed.dev/acp
We use it in our desktop app IDE (mediar.ai) and it's pretty good
Yes it's fully integrated with Claude Skills.
There's first-class integration of https://github.com/enulus/OpenPackage and we provide a ui to install from a list of skills easily as well as add your own.
Dating??? Christ, this planet is cooked.
Surely that is a joke… I hope…
Early on I asked ChatGPT 4 what women actually want. I actually got some advice that was quite helpful.
Alternatively, think about asking the women in your life what they want
While this is generally good advice, it only works if you have women you're close with, at that level, already. If the only women you know are work colleagues, you can't go around asking them for advice on dating (depends on your relationship with them of course, but usually, not work appropriate).
Perhaps that is part of the problem. Talking to women outside of romantic interest might be a good first step
Yes, but that's not useful advice to someone who currently has none.
My point being, maybe other things are foundational to building a romantic life upon. Not saying it is a must but building friendships with all sorts of people will generally help with many aspects of life
Of course. I think that communication is the key to a successful relationship.
However Henry Ford has a well known quote about what people think they want vs what they really want. For that matter, think about how you would answer a question about what you want, vs what you really value to experience in a relationship.
id imagine a skill like "book a restaurant, and update my calendar"
I'm imagining an ADK swipe skill
i use dating skill to update my dating CRM (md notes) and my dating dashboard
recently i have been working on a dating skill so AI book the right place based on each other personality using computer use
please disclose this to the people you’re dating
Please tell me this is a joke.
What's the security boundary here - there's no mention of a VM or anything to isolate the agent from the file system?
i think the point is to access the filesystem. move stuff around write rename etc.
A current concern of mine, also when using Claude Cowork, is that I don’t want files to be modified if I can’t control their versions/ if they aren’t versioned at all.
And even though it doesn’t actually get full sudo access, giving an LLM permission to edit files without being able to track exactly what it’s doing still feels risky.
Still feels a bit technical. The Claude approach is designed for "Susan in Accounting" - what thoughts do you have to reduce the technical barriers?
Yeah it is still too technical.
First obvious stuff like getting the dmg notarized having easier install.
Then after it will be about optimizing onboarding. One of the core goals is to help Susan do 1 small task in under 5 min.
To do that we will need to: - have some prepackaged configs for folks like starter template - ship opencode within the app itself so users don't need to manually install it - and get rid of the technical jargon that is cluttered in the app.
I wonder if most "Susan in Accounting" even know what Claude is. They then have to know to get Claude desktop, get the max plan in order to get Cowork. Most people live on the cloud, so they'll need to get the Claude browser plugin as well to integrate with Cowork.
I'm curious to see if Anthropic can actually win over these non-technical folks as there's still quite a few hurdles for them to get over.
Great stuff and very timely. I just started getting into using opencode and while I'm hugely optimistic about its capabilities and can use it personally without too much sweat, I was left hoping for something a bit more batteries included to give to my non technical colleagues so we can collaborate together. This looks to be exactly what we were looking for so I am looking forward to giving it a spin!
Yeah ! I feel like until we figure out the correct UX for non-technical people the right way would be a sort of hybrid. Where you'd set it up on a remote server (if you know opencode you know openwork) and you just then have non-tech people do a one time setup to connect to the remote and from then on you can easily extend capabilities.
This is the approach that I've taken with Open WebUI. It's a great piece of software for exposing a shared GPT interface but of course that's pretty primitive in the grand scheme of things compared to something like this. But I completely agree with what you're suggesting and I think it's the only practical way to get a multi disciplinary team collaborating with this kind of a tool.
Came up with the same idea this morning and worked half day to get a working version. Then I saw this post.
Will like to contribute.
Created two days ago.... Did you use Claude Code to generate the open source code? :P
I believe Anthropic was saying that they created Claude Work using Claude Code as well, in about a week.
I don't understand the point when opencode desktop already exists.
sounds cool! I'm non technical, I get the workflow with controling home assistant. How could I use this at work? Like how could the workflow be between marketing and devs, for example, for website work, can I change content without dealing with GitHub?
how does it compare to opencode's own gui that uses tauri?
Nice job! How were you able to get it out so quickly?
"Claude build me a Claude Cowork clone, make no mistakes"
yo this is so sick, could def incorporate this into my workflow
let me know if you need help, still a bit rough around the edges
This comment is so fake lol
What is the license?
MIT
god!love you!!
This doesn't seem to be open source, it's currently "all rights reserved" and not under an OSI or FSF license: https://github.com/different-ai/openwork?tab=readme-ov-file#...
Updated ! Thanks for flagging
Awesome! Very cool project, and kudos for making it MIT licensed.
This isn't open source.
Oh damn thanks for flagging. I just added an MIT license.
Beware that the readme.md still says "TBD" at the bottom.
Fixed as well. Thanks again!
The word "open" in software names has completely lost its meaning.
I used "open" because: - it's open source - built on top of open source (opencode) - it's built-around extensibility via plain-text files (skills) and open-source plugins
If everyone put "open" in their project's name because it was open source, almost all the software we use would have "open" in the name
The open source / open plugins / text files is not unique to your project. The majority of the ai tooling space is open in the same way
How much more open do you want it to be? It’s MIT licensed.
100%