No offense, but it seems to me if you really believed in taking updating docs seriously, you'd teach people how to be better at it instead of handing the work off to a gen AI agent and it's potential hallucinatory or swiss cheese responses.
I've worked with dozens of models at this point, and while amazing strides have been made in the LLM space, it's still not at the point where I fully trust it to generate the code AND the docs on how that code works without me diving into every detail and checking for accuracy...and at that point, I might as well just write the docs myself, since it's not really that hard if you have a deep understanding of what is actually happening.
To me, this tool further perpetuates the idea that these "coders" who are relying on gen AI to do the work for them are producing product that they don't know how to fix/update/use because, well, they didn't actually write it. In my line of work, I already spend most of my week fixing issues caused by people who claim they know what they are doing, but don't, so it's hard to not view something like this as future headaches unless it has a high class of proven reliability.
Trust me, I'm happy to be proven wrong. Your landing page isn't really diving into what this thing is doing much, yet you're already asking for a subscription. Explain why we should trust it. What sort of testing was performed to ensure accuracy?