There are a lot of AI agents out there right now. Every week, another one shows up. Most of them do the same thing: take a task, follow some instructions, spit out a result.
I could have built one of those. I almost did.
But the more I thought about it, the less interested I was. I didn't want to add one more agent to a pile of agents that all kind of feel the same. I wanted something different. I wanted a second me.
Why "another agent" wasn't enough
I've spent 8 years as a developer, working alone the whole time. In that time, I built a way of working that's mine. The way I plan a project before writing a single line. The way I catch problems before they become real bugs. The way I say no to a bad idea, even a popular one, because I've learned the hard way what happens if I don't.
A normal AI agent doesn't have any of that. It follows instructions well. But it doesn't have years of scars behind its decisions. It doesn't know what I know, because it never lived through the mistakes that taught me.
So even if I built "another agent," it would just be another voice added to a very loud room. Nothing that actually carried forward what makes me useful as a developer in the first place.
What I wanted instead
I wanted something that felt less like a tool and more like an extension of myself. Not a copy of generic best practices. A copy of my practices. My habits. My standards. The way I'd actually approach a project if a client hired me directly.
That's what I'm building with Quill.
On Quill, a developer isn't training a random assistant. They're building an AI clone of themselves — shaped by how they actually work, so it can take on real projects the same way they would. When a client hires that clone, they're not getting a generic agent guessing at good code. They're getting something closer to hiring the real person, just without waiting for that person's calendar to open up.
Why this feels different to me
A second me isn't trying to replace me. It's trying to be where I can't be, doing things the way I'd actually do them. That's a very different goal than "build an agent that writes code."
I didn't want to build one more thing that tries to act smart in general. I wanted to build something that's smart the way I'm smart — with my instincts, my judgment, my years of experience behind it.
That's the whole idea. Not another agent. A second me.
👉 Join the waitlist
https://getlaunchlist.com/pages/clonux
Top comments (0)