Xengine Enhancements & AI Inbox Autopilot Launch — July 22, 2026

Today was a long 20-hour grind, but I made significant strides on my journey towards building a one-man-show company with a billion-dollar valuation. I shipped 16 commits, focusing primarily on enhancements to the xengine, the launch of the AI Inbox Autopilot, and various improvements to the directory features. The key highlights include a new posts-per-day dashboard, a marketing page for the AI Inbox Autopilot, and the establishment of a microblogging network.

The centerpiece of today's work was the introduction of the posts-per-day dashboard dial on the xengine. This feature allows users to vary their posting schedule with a simple header stepper interface. I implemented a new control route that enables the configuration of posts per day, which can be clamped between 1 and 24. This involved a lot of back-and-forth with the Mac engine to ensure that it operated smoothly, but the end result is a more user-friendly experience that encourages engagement.

In conjunction with this, I also rolled out the AI Inbox Autopilot marketing page. This feature is pivotal for my vision of the agentic web, as it allows users to interact with their Gmail accounts in a streamlined manner. The page includes a checkout option priced at $49/month, which I believe will attract a solid user base. However, creating the middleware to serve this page without redirecting to a portal slug proved to be a challenge, requiring some late-night debugging to ensure everything worked seamlessly.

Privacy compliance was another critical area I focused on today. I added the necessary disclosures for Google user data and Gmail Limited Use, which are essential for the AI Inbox Autopilot's functionality. This involved updating the policy language and ensuring that users have control over their data, which is non-negotiable in today's landscape. The process of integrating these disclosures was time-consuming, but necessary to build trust with users.

I also made substantial improvements to the directory features, including the addition of a microblogging network. The new Lane-A adapters allow for a more robust interaction within the network, enabling users to share their posts across various platforms without friction. However, I faced a few hurdles while verifying the functionality of these new adapters, which ate up a few hours of my day. The issues stemmed from the need to ensure that each adapter worked correctly with the existing infrastructure, but after multiple tests and tweaks, I finally got it stable.

To further enhance the directory experience, I implemented an hourly autopilot scheduler, which ensures that queued listings are processed correctly. This feature was critical, as it allows for a more reliable drip of content without overwhelming the system. I encountered some bugs while configuring this scheduler, particularly with the gating mechanisms, but I managed to resolve them after a few iterations.

I also had to address some bugs in the admin navigation. Two commits were dedicated to restoring the missing Directories link in the Growth navigation. This issue arose from a stale-checkout clobber that wiped the directory engine, and it took longer than I anticipated to trace the problem and restore the link.

Reflecting on today’s work, it’s clear that building this platform solo with the help of AI tools like Claude Code in VS Code has its ups and downs. While it allows for incredible flexibility and control over the development process, it also means that every bug and every feature is mine to tackle alone. Today was no exception, as I spent a significant amount of time troubleshooting and refining features that will ultimately enhance the user experience. I’m proud of what I accomplished, but I know there are still many challenges ahead as I strive to create a billion-dollar valuation company with just myself and AI as my team.