alice alexandra moore

How many times have you explained the same login flow to your coding agent?

We kept doing that with every bug, so we turned the whole routine into skills. Now an agent picks up a bug report, reproduces it, fixes it, and hands us a tested PR to review.

Here's how to set it up yourself, starting with one bug.

builder.io/blog/build-an-…

the big walk stuff has me sad.

shaming artists for ai use doesn't make less ai use happen.

it just makes more secret ai use happen.

i want to live in a world where people are excited about and talk about how they make things.

because then we get to learn from their processes, improve those processes, and make cooler things.

when the process itself is hurting people, that's something we should be talking about and fixing, too. but the answer is never just cancelling it all or pretending it's not happening.

🔥 $10,000 in prizes. Agent-Native Hackathon. 👾

Most AI agents still hand you a wall of text and make you go dig for what they actually changed.

So we're running a hackathon for agents with actual interfaces. Build an app where the agent has a real job, and people can see its work, edit it, and tell it what to do next.

That's how I run my own task list now. I leave a comment on a task, an agent does the work, and I fix whatever it got wrong right there on the page.

It's free, open worldwide, and you can go solo or bring a team. Submissions close Nov 9. Agent-Native is free and open source, so starting is one command.

Please stop handing your AI agents your API keys.

I know, I'm guilty too. My agents run as me, with my GitHub token and the keys in my .env files.

But there are better ways now. I'm slowly switching over.

I wrote up what they are, and why attackers spent October going after keys like mine.

builder.io/blog/ai-agent-…