jdeng

Good comparison of the 2 approaches of agent harness. Vita AI[0] also picked the outside model. We used the E2B sandbox solution, and fortunately it supports volume, so we just create volumes for personal skills and organization workspace, and mount them to the sandbox. This elegantly solves the "file system" problem you mentioned.

[0]: https://www.vita-ai.net (AI agent for content creation & social media management)

Excited that the long awaited v4 is finally out. But feel sad that it's not multimodal native.

For openclaw to become helpful, you have to connect it to your personal email, access to your file etc. All of these requires user's manual setup right?. I do not get the point of "batteries included". Installing it is not the bottleneck right? The official docs has detail procedures for all deployment options.

Good try. I had the idea before, but later realized all the agent ecosystem has already highly coupled with bash scripts including skills. Though bash is not a good programming language, it's indeed flexible for CLI apps. I also realized that what agent and skills really need is just a glue language, not a real programming language, so bash is not so bad in the case.

Working on a Vibe Testing Agent (https://www.vita-ai.net/), a perfect companion of vibe coding. Now the loop is vibe coding <-> vibe testing.

Glad to to see open source models are catching up and treat vision as first-class citizen (a.k.a native multimodal agentic model). GLM and Qwen models takes different approach, by having a base model and a vision variant (glm-4.6 vs glm-4.6v).

I guess after Kimi K2.5, other vendors are going to the same route?

Can't wait to see how this model performs on computer automation use cases like VITA AI Coworker.

https://www.vita-ai.net/

Exciting to see Anthropic validate the "AI coworker" direction. We're building VITA AI (https://vita-ai.net) with similar philosophy but for enterprise QA testing.

One key architectural difference: Cowork runs sandboxed VMs on your local macOS machine, but we run sandboxes entirely in the cloud. This means:

- True isolation - agents never touch your local files or network, addressing the security concerns raised in this thread

- Actual autonomy - close your laptop, agent keeps working. Like delegating to a real coworker, not pairing with an assistant

- Scale - spin up 10 test agents without melting your CPU

The trade-off is latency and offline capability, but for testing workflows (our focus), asynchronous cloud execution is actually the desired model. You assign "test the checkout flow," go to lunch, come back to a full test report + artifacts.

Different use cases, different architectures. But the broader trend feels right - moving from conversational assistants to autonomous agents that operate independently.

Sure. Join the waitlist below and I will note down that you need private demo. I will contact you once we got bandwidth on that (private demo).

https://www.vita-ai.net/waitlist

Working on an AI coworker (QA engineer) that can autonomously tests web app like your human QA colleagues. It has its own virtual desktop and runs as background agent.

https://www.vita-ai.net/

Thanks! Voice input and AI assisted notes really save me. Whenever I have some spark, I just talk to AI and ask it to transform my thoughts into notes.

>A bow is more powerful than a stick but also develops more motor skills, a bike develops the users body more than a car, and so on.

Totally agree. I like bike much more than car, however the trend of bike superseded by car cannot be reversed, given car's convenience and power.

IMO if we cannot reverse it, then embrace it. It applies to other tech advancement as well.

The decline of deep thinking in the digital age is a pressing concern. While technology presents challenges to our cognitive faculties, it also offers unprecedented opportunities. By leveraging AI as a cognitive partner, we can counteract the negative effects of constant connectivity and revive our capacity for deep, meaningful thought.

Thanks for the advice. Since the conversation itself is long, integrating it into the blog would dilute the article. I will try to make the link more accessible, and add a TL;DR section.

There is a ton of curated list of awesome stuffs, as you can explore on http://awesome.re. But I found it inconvenient to search by topics, or preview projects while browsing. Do you feel the same? Is the pain point serious enough for a search engine of curated awesome stuffs?

I was searching for a usable/scalable way to develop cross platform native mobile app, but found limitations of current solutions.

StateX (https://github.com/lilac/statex) is a cross platform native application architecture.

How do I move this to "show HN" list?