yousefh409

Thanks for the feedback! The $49 of usage is used across AI inference, sandbox minutes, and so on. We mention it later on in the pricing page, but will make sure it is more clear fromt he onset.

And yes, the LLM safety judge is not in general access right now. Still experimenting, and it is not being used by any customers currently.

Great question! We have a couple things: (1) We have strict gaurdrails that makes sure that all data displayed in the generated comoponents is accurate and sourced from a real API call (2) We learn your design system and brand, so that everything Vendo generates looks and feels native (3) We have our library of pre-built components purpose-built for Vendo and how it works, which helps us increase the quality even more

Great points!

Great points, and this is a problem we have spent a lot of time thinking about as well. Our current hypothesis, is that the models will reach (and have already arguably come very close) a level where problems like these can be resolved without any human intervention needed. That doesn't mean that problems won't arise, they for sure will (nothing will ever get to 100%), but the increasing strength of the agents will allow resolving those issues to become very consumer-friendly.

I would also encourage you to take a look at our benchmarks (linked in the post), that show how the gaurdrails we have in-place allow us to deliver pretty strong results to our users. And as a last point, most of our users are not creating business-critical features with Vendo right now. For a lot of them, it is an additional feature that they wished they had but didn't. And for the companies we work with, the benefit of providing their customers with this level of customization outweigths the possible risk that you mention.

Would love to hear more of your thoughts, thanks for the feedback!

That's really interesting. Would love to learn more about what type of customizations/features you have seen your users create?

As detailed as the user wants. We don't limit the user with things like pre-built components and so on, so they can let their creativity run wild. What we chose to focus on instead, is making sure that the things they create are sandboxed, accurately representing the data at hand, and feel native in the product.

Great question! Honestly, not a lot. This is moreso for the rest of the population that does not even know Claude Code exists. A lot of our customers are vertical SaaS's that have tons of non-technical users using their product every day, and we want to bring them the power of claude code right inside their product.

I think you bring up a good point though on how consumption of software will evolve (and the possibility of Claude just being the entrance to everything), and I think this is soemthing we have been thinking a lot about as well.

Why would any company leave the enforcement of rules to an agent? If something is truly a rule, there should be code that deterministically enforces it.

That is really interesting!

Agreed that agent harnesses as the primary interface is also a big possibility. From our conversations though, a lot of companies also don't want to lose complete control over the user experience. Which is why I think an increasingly more likley direction is every company having their own agent that functions as a "personal concierge" that gets stuff done for their usrs user and gives them what they want, but all under their own control.

I think it is a really interesting space in general, and looking forward to continue building in it!

Would also love to hear what others have been thinking about in this space, as well as any feedback/criticism!

Really cool! Hermes and openclaw setup was way too hard, this is much needed