Fresh/Brewed

AI Skills with Cloudflare+Ngrok Tunnels

I might have spent too much time on the demo...

In my last article I spoke on CLoudflare tunnels as well as Ngrok ones. As we were wrapping that, I found a clean approach to just adding a sidecar to a Kubernetes pod to create tunnels on the fly. I was using a Wilderness app demo to do that but thought it might be something I would want to do again (and of course share with others) Creating an AI Skill Using Antigravity with Argon 4, I started by getting on the right branch and asking it to make a Skill for the helm chart:

Cloudflare and Ngrok Tunnels

How is this any different than ngrok?

I recently came across this Cloudflare blog article on Quick Tunnels. They seem like lightweight ngrok tunnels, but without needing any auth. The article suggested they were great to use with agentic flows, but I have a hard time these are things we get just for free - no account required. We’ll look into Cloudflared Tunnels on the CLI, how we might set them up in Kubernetes then compare these things to what Ngrok has to offer.

Gemini 4 and Rust TUIs

Ratatui with a bit of Argon

We last looked at Ratatui and VCR back in July. I recently had a need for VCR and it reminded me that Ratatui was something I demo’ed but never really tried to use. With Google Gemini 4 Argon releasing today, I thought this might be a fine opportunity to test it with some new and rust based. I’m not just going to make a TUI that does nothing. Instead, I’ll be pairing this with a demo I’ll be giving down in Rochester MN for the Code and Connect night.

OS Apps: Timezone, Swiish and VaultWarden

Time, Cards and Secrets

I have a few apps on my todo list today. The first of which is this timezone app in Linux. I’ll try a few ways to install. The next is Swiish which is an excellent containerized business card app. Lastly, we’ll look at Bitwarden and Vaultwarden as I try to restore my OS secrets engine on the cluster. But let’s start with a simple timezone app… Timezone I found this timezone app some time back and really wanted to get it a try.

A2A

Bring me my Llama

Ever since I heard about the Agent to Agent (A2A) protocol, I’ve been curious what it was and how it might be useful. The general idea of A2A is that it’s a standard that Agentic workloads can leverage to discover and talk to each other using “cards” that advertise capabilities. We will look at two A2A examples, how they work and test them. Then we’ll try and recreate them with python containerize servers with REST endpoints.