Back in the Lab: Goats, Control Planes, and the Future of Small ISPs

I'm back in the lab after a few weeks of downtime—life, work, and the occasional mandatory yard battle. CodexMCP is alive again, and with it come the projects it has spawned like some sort of digital hydra with a horned fetish.

Yes, everything is named after a goat. No, it’s not a coincidence. CodexMCP started as the Master Control Program for managing infrastructure sanity. Then came GoatMUX, the SIP routing and monitoring brain. Then GoatWatch, the log-watching and anomaly-detecting sentry. Why goats? Because they eat garbage and thrive on mountainsides, just like the infrastructure I inherited. It’s also funny, and if you’ve ever tried to trace SIP call failures between two beige plastic boxes from 2007, you know humor is a survival mechanism.

But let’s get serious for a moment.

These aren’t just pet projects. They’re the prototype nervous system for running an entire small to mid-sized ISP—with thousands of users—on a skeleton crew. I'm talking two or three people max. No bloated vendor stacks, dependency hell, or million-dollar "support" contracts that forward you to tier zero in Bangalore.

Every tool in this stack is open source. Every decision is made with operability in mind. CodexMCP is not the tool—it is the coordination layer. It’s the brainstem. It knows what tools to call, how to call them, and when to escalate. It’s not just a dashboard. It’s not just automation. It’s an intelligence overlay that ties RADIUS, DHCP, SIP, DNS, NAT, logging, and user activity into a single operational model.

That’s the endgame.

Not a monitoring system.

Not a provisioning platform.

An autonomous infrastructure command system for ISPs. Designed for small teams. Built from the ground up for observability, survivability, and sanity.

I’m currently working through DHCP simulation (15,000+ virtual clients in a testbed), refining log analytics through GoatWatch, and prepping the next iteration of SIP-aware routing under GoatMUX. The ideas list grows by the day—some are serious, some are absurd, but all of them aim at reducing human overhead without giving up operational control.

I’ll be posting more frequently now, not for attention but as documentation, for myself and for anyone else stuck holding together legacy telco infrastructure with duct tape and sarcasm.

Stay tuned if you’re interested in seeing where this goes. If you’re an ISP operator tired of paying six figures for broken tools, I’ll eventually have something for you, too.

-Breaks are Good
--Bryan