Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

tilde.run

Post reply on HN
Print view View original

91–100 of 147 posts

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#91

Post by messh »

docheinestages wrote: post #10

Just my two cents: less is more and the first impression matters a lot. I'm saying this because we see a new agent sandbox tool on the front-page almost every day. Most of them have an AI-made landing page design, lots of animations, lots of words. This has become a bad sign for me. I can tell that you put time into it, made a video, and everything, but I guess I'm suffering from some kind of fatigue of having to go…

Sadly this is what sells. Standing out in this regard checkout https://shellbox.dev maybe swinging too far though?

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#92

Post by nateb2022 »

doctorpangloss wrote: post #78
ozkatz wrote: post #13

Earlier quoted context omitted.

That's fair. What makes this unique is the versioned, composable filesystem. It's built on top of lakeFS ( https://github.com/treeverse/lakeFS ) so it scales really well, unlike other solutions that try and do this with Git directly.

LLM authored comments are against the rules. I don't think file versioning is differentiated anyway.

OP is actually one of the co-creators of lakeFS, for context.

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#93

Post by chickensong »

gverrilla wrote: post #69

I'm far from an expert on the field or in computer science, but from my limited perspective I don't see the need for sandboxing - after thousands of claude code interactions it never did nothing wrong that was serious, at all. If I understand this all correctly, lakeFS would be useful for versioning huge dataloads - but it's not my case: for my usecase I use dura and that's plenty, and for more serious projects where…

You're basically saying there's no need to wear a seatbelt because you've driven thousands of times without an accident. Claude is pretty well behaved, but it's not guaranteed to be safe, especially as you start to hit the gas and relinquish more control. Hope for the best, but plan for the worst and all that. Just because your use case doesn't need sandboxing, doesn't mean there's no need for sandboxing.

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#94

Post by viewhub »

ozkatz wrote: post #32
viewhub wrote: post #29

What compute resources does the sandbox have? Memory/CPU/GPU?

Currently a static 2 cores and 4GB RAM, no GPU. Will be configurable soon!

Cool. I'll take the API for a spin in the next week. If I use it for my upcoming project, I'll need the ability to control the available CPU/GPU/Memory attached to each sandbox so I can right-size it for the workload. Congrats on the launch!

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#95

Post by arm32 »

sahil-shubham wrote: post #76

Nice work on the website! Building something for the same problem but more so from the perspective of self-hostable stateful sandboxes, and not just the filesystem (see https://bhatti.sh ). What sandbox solution are you using here?

Your landing page looks very similar to OP's! I thought it was the same site!

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#97

Post by gverrilla »

chickensong wrote: post #93
gverrilla wrote: post #69

I'm far from an expert on the field or in computer science, but from my limited perspective I don't see the need for sandboxing - after thousands of claude code interactions it never did nothing wrong that was serious, at all. If I understand this all correctly, lakeFS would be useful for versioning huge dataloads - but it's not my case: for my usecase I use dura and that's plenty, and for more serious projects where…

You're basically saying there's no need to wear a seatbelt because you've driven thousands of times without an accident. Claude is pretty well behaved, but it's not guaranteed to be safe, especially as you start to hit the gas and relinquish more control. Hope for the best, but plan for the worst and all that. Just because your use case doesn't need sandboxing, doesn't mean there's no need for sandboxing.

I'm not having a debate because I'm quite ignorant of the subject. Just trying to learn from you: wouldn't recoverability and observability suffice instead of sandboxing, if such events are indeed rare? not necessarily for all usecases, but for most?

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#98

Post by ozkatz »

gatvol wrote: post #87
_pdp_ wrote: post #84

Git is already versioned, S3 support versioning and any file copied into the sandbox, is well a copy, so I am not sure what is the angle here. Other than that it looks cool!

Doesn't s3 now have versioning + POSIX mounts?

S3 offers versioning at the single file level.

Imagine an agent dropping a directory with 1m images in it. just figuring out what happened and what got dropped, restoring it one by one, etc. - doable, but ergonomics are a bit lacking.

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#99

Post by ozkatz »

mehmetkeremmtl wrote: post #80

The versioned filesystem is exactly what's missing when agents hallucinate and go off the rails. How fast are the rollbacks if an agent completely messes up the directory state?

very very fast: proportional to the count of objects modified, but not their size. Every commit represents a snapshot - an immutable listing of objects that represents the repository. reverting is essentially applying the inverse of the diff introduced by the reverted commit.

This is metadata only as the objects themselves are immutable.

Re: Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem

#100

Post by ozkatz »

whalesalad wrote: post #79
ozkatz wrote: post #23

Earlier quoted context omitted.

https://www.youtube.com/watch?v=fDR8tmes020 - a 2 minute hands-on demo!

Being brutally honest - terrible demo. 80% of this is baseline stuff, setting up permissions (annoying), and the last few seconds we see a file was deleted and we can approve it. This is not selling your product.

Appreciate the honest feedback. I agree there's a lot to improve there.
Post reply on HN