notgelotto

Those are the current things we are wrapping up before the MVP release. We don't want to just make it easy to deploy, scale, & secure, we want to give you full insight into how your service is doing.

We’re experienced engineers who’ve worked on large-scale distributed systems. We’ve been using Claude heavily to help with architecture decisions, code design, testing strategies, and rapid iteration on complex infrastructure.

The result is Boogy, prompt it (or write Rust) to generate full backends with an embedded high-perf DB (faster than SQLite on mixed workloads), vector search, auth, and durable jobs. One curl to deploy. Services call each other in-process for microsecond latency.

We’re planning to open it up soon and make it completely free so people can properly battle test it.

Would love to learn about what issues current vibe coders might be running into in production?

I kept starting Claude Code sessions on my dev machine, leaving my desk, and having no way to check on them from my phone.

So I built HQSSH — a mobile SSH app with a small Go daemon (hqsshd) that runs on your remote systems. It discovers your git repos, detects which AI tools are installed, and lets you launch them in the right directory with one tap. Sessions persist across disconnects and devices.

Everything goes over SSH — no cloud relay, no account, no extra ports.

Daemon is open source: https://github.com/Gelotto/hqsshd