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

Claws are now a new layer on top of LLM agents

twitter.com

Post reply on HN
Print view View original

421–430 of 1001 posts

Re: Claws are now a new layer on top of LLM agents

#421

Post by qup »

brap wrote: post #309
zozbot234 wrote: post #54

Earlier quoted context omitted.

You could run them in a container and put access to highly sensitive personal data behind a "function" that requires a human-in-the-loop for every subsequent interaction. E.g. the access might happen in a "subagent" whose context gets wiped out afterwards, except for a sanitized response that the human can verify. There might be similar safeguards for posting to external services, which might require direct confirmat…

So you give it approval to the secret once, how can you be sure it wasn’t sent someplace else / persisted somehow for future sessions? Say you gave it access to Gmail for the sole purpose of emailing your mom. Are you sure the email it sent didn’t contain a hidden pixel from totally-harmless-site.com/your-token-here.gif?

I don't have one yet, but I would just give it access to function calling for things like communication.

Then I can surveil and route the messages at my own discretion.

If I gave it access to email my mom (I did this with an assistant I built after chatgpt launch, actually), I would actually be giving it access to a function I wrote that results in an email.

The function can handle the data anyway it pleases, like for instance stripping HTML

Re: Claws are now a new layer on top of LLM agents

#422

Post by orsorna »

qup wrote: post #407
xg15 wrote: post #51

Earlier quoted context omitted.

What exactly are they self hosting here? Probably not the model, right? So just the harness? That does sound like the worst of both worlds: You get the dependency and data protection issues of a cloud solution, but you also have to maintain a home server to keep the agent running on?

"maintain a home server" in this case roughly means "park a headless Mac mini (or laptop or RPi) on your desk" And you can use a local LLM if you want to eliminate the cloud dependency.

You have spend tens of thousands of dollars on hardware to approach the reasoning and tool call levels of SOTA models...so, casually mentioning "just use local LLM" is out of reach for the common man.

Re: Claws are now a new layer on top of LLM agents

#423

Post by qup »

cryptoegorophy wrote: post #391
unixfg wrote: post #350

Earlier quoted context omitted.

My favorite use so far has been giving it a copy of my Calibre library. After having it write a few scripts and a skill, I can ask it questions about any book I’m reading. This week I had it order a series internally chronological. I could use the search on my Kindle or open Calibre myself, but a Signal message is much faster when it’s already got the SQLite file right there.

I am sorry to sound dumb but can’t cursor ai do this same thing? They have .md files with skills and knowledge

What's the relevance?

Re: Claws are now a new layer on top of LLM agents

#424

Post by squidbeak »

YeGoblynQueenne wrote: post #276
mattlondon wrote: post #31

Earlier quoted context omitted.

I think for me it is an agent that runs on some schedule, checks some sort of inbox (or not) and does things based on that. Optionally it has all of your credentials for email, PayPal, whatever so that it can do things on your behalf. Basically cron-for-agents. Before we had to go prompt an agent to do something right now but this allows them to be async, with more of a YOLO-outlook on permissions to use your creds,…

I think this is absolute madness. I disabled most of Windows' scheduled tasks because I don't want automation messing up my system, and now I'm supposed to let LLM agents go wild on my data? That's just insane. Insanity. Edit: I mean, it's hard to believe that people who consider themselves as being tech savvy (as I assume most HN users do, I mean it's "Hacker" news) are fine with that sort of thing. What is a person…

> and now I'm supposed to let LLM agents go wild on my data?

Who is forcing you to do that?

The people you are amazed by know their own minds and understand the risks.

Re: Claws are now a new layer on top of LLM agents

#425

Post by dragonwriter »

jstummbillig wrote: post #406
thegrim33 wrote: post #394

Earlier quoted context omitted.

How do you need to supervise this "less" than an LLM that you can feed input to and get output back from? What does it mean that it's "running continuously"? Isn't it just waiting for input from different sources and responding to it? As the person you're replying to feels, I just don't understand. All the descriptions are just random cool sounding words/phrases strung together but none of it actually providing any c…

> Isn't it just waiting for input from different sources and responding to it? Well, yes. "Just" that. Only that this is at a high level a good description of how all humans do anything, so, you know.

Yeah, and if you give another human access to all your private information and accounts, they need lots of supervision, too; history is replete with examples demonstrating this.

Re: Claws are now a new layer on top of LLM agents

#426

Post by cap11235 »

cryptoegorophy wrote: post #391
unixfg wrote: post #350

Earlier quoted context omitted.

My favorite use so far has been giving it a copy of my Calibre library. After having it write a few scripts and a skill, I can ask it questions about any book I’m reading. This week I had it order a series internally chronological. I could use the search on my Kindle or open Calibre myself, but a Signal message is much faster when it’s already got the SQLite file right there.

I am sorry to sound dumb but can’t cursor ai do this same thing? They have .md files with skills and knowledge

I'd imagine you could (never used Cursor myself though). I do a similar thing with my collection of books, but I just use Claude Code.

Re: Claws are now a new layer on top of LLM agents

#427

Post by AlecSchueler »

BeetleB wrote: post #360
krelian wrote: post #127

Earlier quoted context omitted.

Maybe I'm missing something obvious but, being contained and only having access to specific credentials is all nice and well but there is still an agent that orchestrates between the containers that has access to everything with one level of indirection.

I don't see why you think there is. Put Openclaw on a locked down VM. Don't put anything you're not willing to lose on that VM.

But if we're talking about optionally giving it access to your email, PayPal etc and a "YOLO-outlook on permissions to use your creds" then the VM itself doesn't matter so much as what it can access off site.

Re: Claws are now a new layer on top of LLM agents

#428

Post by luckylion »

throwaway13337 wrote: post #249

The real big deal about 'claws' in that they're agents oriented around the user. The kind of AI everyone hates is the stuff that is built into products. This is AI representing the company. It's a foreign invader in your space. Claws are owned by you and are custom to you. You even name them. It's the difference between R2D2 and a robot clone trying to sell you shit. (I'm aware that the llms themselves aren't local b…

It always depends on who you consider the user. The one who initiated the agent, or the one who interacts with it? Is the latter a user or a victim?

Re: Claws are now a new layer on top of LLM agents

#430

Post by alecco »

> Bought a new Mac mini to properly tinker with claws over the weekend.

Disappointing. There is a Rust-based assistant that can run comfortably in a Raspberry PI (or some very old computer you are not using) https://zeroclawlabs.ai/ https://github.com/zeroclaw-labs/zeroclaw (Built by Harvard and MIT students, looks like)

EDIT: sorry top Google result led to a fake ZeroClaw!

Post reply on HN