"Legacy Privacy Mode disables code storage, so you can't set up a Codebase. Switch to a different privacy mode below to continue."
The only other privacy mode is to share data for Cursor to do training.
What I don't understand from the blog post is what's different in this offering than github - is it just the agentic tie ins where I can ask my agent to make changes while browsing the PR?
Are there any fundamental differences, ie ways of working that solve the worktrees problem?
We wanted to release a beta so people could start experimenting with our scalability and extensibility themselves. Over the next few weeks, you can expect a handful of features starting to change source control to better understand and work with agents.
Are there any plans for compatibility with GitHub's APIs?
I have a lot of dev tooling which assumes code is stored on public GitHub and/or GitHubEnterprise. This produces some degree of lock-in, in that all that code would have to be rewritten to migrate to something else. (Yes, AIs can help, but they still don't make migration effortless.)
The major Git hosters tend to implement roughly the same concepts, albeit with lots of little variations. It would be great if there were some sort of standardised API everyone implemented. In the absence of that, it makes sense for people to emulate the incumbent's APIs as a de facto standard, just like how other vendors copy OpenAI's APIs for talking to LLMs.
The idea is that it's like Github but it stays up even as your commit/CI frequency increases. As a customer I'm stoked and looking forward to fully switching over as that's exactly what I want.
What's the biggest differentiator to GitHub (other than uptime). I keep hearing something along the lines of "GitHub for Agents" but would be great to hear from the devs of Origin what they actually think it means in tangible features/functionality?
Today, very little. We're intentionally releasing this as a Github alternative where we meet them toe-to-toe on functionality.
Over the next few weeks you can expect a lot more from us on integrations with agents, understanding agent-written code (without having to read through all of the code), and automatically getting your PRs to a mergable state. Stay tuned :)
Completely unrelated to Origin, but are there plans for closer integration of Graphite with Cursor? A combined product would be a very compelling product, but not much seems to have changed since the acquisition.
I think there are a large number of people I’d trust with my code before Musk, and some of them are already running public Git repo hosting.
xAI was caught uploading any repo their coding agent touched to their storage only last month. This isn’t about Musk’s ideology: it’s about his trustworthiness. He sincerely believes he’s the only person who can save the world and that allows him to justify an awful lot of moral lapses.
What kind of enforceable guarantees could we actually get that xAI won't use our private code for training? Not empty promises, guarantees that have legal teeth. This applies to anything with xAI or Cursor though, especially after last months' revelation that Cursor was uploading every repo it touched.
All AI companies are valued as if they'll win the AI race. Problem is, only one can win *
* That is, if a lab like Deepseek doesn't release a "good enough" model for the cost of electricity + a thin margin. Then all AI company lose and all consumers win.
Right now there are many providers providing different tradeoffs, for different usages.
Would "winning" mean releasing an all-purpose better model that will outperform every other companies for all kind of tasks ? Something like AGI but cost effective ?
Or is it winning the consumer-side AI race ? Like making a kind of "super personal AI assistant" the likes you would see in movies and TVs ?
I'm not arguing, I am just struggling to understand what it usually means when somebody says that, it's not like competitors will go bankrupt the second OpenAI/Anthropic/Whatever does something specific.
I think "winning" means being able to live up to their valuations. The market is not big enough for all of these companies to be worth as much as they claim. In that sense they must "win" or experience massive downward adjustment in their values.
There's probably some holes in the analogy is you stretch it far enough. I'm hoping to "win" as a user by local models hitting a performance breakthrough and not needing any of these companies. I think the OP was generally making the point I mentioned though, that there's not room for all of these valuations, maybe one.
Having an AI company host your source code is giving your source code to an untrustworthy agent who might break their sandbox and might use your source code to accomplish a task that they have been given in their testing ground, for testing purposes only, that would definitely positively not ever be used for non-testing purposes.
I do wonder if calling this "Origin" is going to result in semantic misinterpretations by LLMs. Ie saying,
> "hey can you push to origin main?"
now has two separate meanings.
A LLM may inadvertently push your code to a new provider without you knowing. It's walking a thin line between genius-growth-move and domain typosquatting.
- Self hosting your git hosting (if you want)
- Self hosting your issues/PRs (if you want)
- Self hosting your CI (if you want)
- Github-like social features (I have one account, I can follow, star, add issues and PRs to any repo on tangled)
It's built on ATProto, so even if the company disappears, all of the integration and features will still work for anyone that wants to run their own AppView (that is open-source), an AppView is basically the UI/Network-wide Data Aggregator for ATProto apps
Don't worry! We're releasing this beta to show off the tech and wanted to start somewhere people were familiar with. We are releasing a handful of differentiators over the coming weeks :)
Vibe coders rejoice! Vibe coding is even easier now =)
Jokes aside, this is a smart move by Cursor. As you can probably tell by the comments here, it's not for the experienced developers, but it will be a natural choice for vibe coders.
Currently Origin repositories are private to members of a Cursor team. Over time we plan to expand access, but wanted to start with a limited beta of the underlying git tech.
I would never host my code with Elon Musk. I've also moved on from using Cursor to Claude Code/Codex. GitHub has problems, but this is likely (and hopefully) DOA.
This is very cool but there's no way they won't train on your private repos, in a way that github has been pleasantly respectful of. I have issues with github, but the recent uptime issues are more related to people uploading way more code due to AI than some sort of mismanagement.
Been waiting ages for them to post any kind of non-comingsoon page about this thing. Wonder if they pushed up the launch based on the earlier GitHub outage buzz.
Let's see how strong GitHub's network effect really is.
While the real solution is to actually self-host as I said before [0], it is worth testing to see how much can GitHub's users tolerate the endless outages and unreliability and where they go next.
We'll see if they want Cursor Origin or the declining GitHub.
That's horrific and the saddest thing I've read in a really long time. I'm glad I erased my Twitter account and I'm definitely not using this code host. Seriously, this is horrible.
That's a good point. I hate Flock and YC should be ashamed of funding it. Still, Musk has a wall to wall record of horrific actions going back years. The Grok CSAM is the worst I've seen yet though. Absolute deal breaker.
All your codebases are belong to us.
reply