I built an AFK coding-agent loop so I could stop babysitting agents at 4am.

I spent most of my career doing non-technical work. Now I build our product. The bridge was months of late nights talking to AI coding agents, hitting approve every thirty seconds until I gave up and went to bed.

So I wrote up how I stopped: blog.heynews.co/afk-coding-agent-loop

The short version: I wanted my nights back. Instead of sitting next to an agent like a nervous parent, I built a queue I fill during the day and walk away from.

I write GitHub issues with enough context that an agent can finish one without coming back to me for a product decision. The ready ones get a label. From there, the agents carry each issue through implementation, review, fixes, merge, and deploy checks, then unblock the next one. I stay in charge of what gets built. They do the rest.

A few things that made it work as a founder, not just a demo:

• the decision of what to build stays human; only the labor moves to the agent
• labels as the state machine, so I can read the status of any work at a glance
• review runs as a separate loop with fresh context, not the same agent that wrote the code and is now sure it's right
• a host-side gate that checks GitHub directly instead of trusting the agent's "I handled it"
• nothing reaches staging without a PR, ever

The real lesson wasn't about code. Getting an agent to write code was the easy part a year ago. The hard part is the structure around it: enough that the work stays legible and safe to leave running while you're not watching. That structure is the moat, not the model.

If you're building solo or running a small team, where do you keep yourself in the loop versus letting the system run, and what's the first repetitive part of your business you'd hand to a queue like this?