The Plugin I Wanted Didn't Exist. So I Built It.

For most of my career there was a clear line between "3D artist" and "tool developer." If the software I used every day was missing a feature, or doing something in a way that cost me time on every project, my options were limited. I could wait for the vendor to ship it, which usually meant years. I could post on a forum and hope somebody with a CS degree and free time built it for fun. Or I could try to learn Python deep enough to write the thing myself, which for most working artists meant "not going to happen."

That line has moved.

The old way

Back in 2019 I wrote my first Modo plugin. It was small. It took weeks. I had a clear picture in my head of what I wanted it to do, and maybe 80% of my time went into fighting the language, the API, and the scripting docs to make that picture real. The other 20% went into the part that actually mattered: how the tool felt, how it flowed, whether it saved me time or just moved the friction somewhere else.

That ratio is not a good ratio when your job is 3D, not code.

I kept at it. Over the years I built a small collection of Modo tools. Each one shaped itself around how I actually worked, because I was the one who had to use them. Each one took longer than it should have. And each one taught me something I'd have to relearn the next time I opened a code editor after six months away.

Then Modo was discontinued

If you've never had the software you built your craft around walked into the sunset, I can tell you what it feels like. A shelf of custom tools you made with your own hands, slowly becoming unusable. Years of work, still functional for now, with an expiration date attached.

The honest, non-AI version of the next chapter was grim. Port everything to Blender. A different scripting language, a different API, a different UI paradigm. Learn a new system from scratch while also doing my day job. Best case, a year of nights and weekends. Worst case, give up on the tools entirely and live with stock software like most artists do.

What Claude Code actually changed

Instead, I started working with Claude Code to port my plugins over to Blender. And I want to be specific here, because there's a lot of noise around AI tooling right now and not enough plain talk.

Claude Code saved me hundreds of hours. That is not marketing language. That is a real count of the days I would have spent reading API docs, debugging stack traces, and translating Modo idioms into Blender idioms by hand.

The thing that actually mattered, though, wasn't the hours. It was what the hours were freed up for. I got to focus on the parts that had always gotten 20% of my attention. How the tool feels. How the UI reads. Where the shortcuts live. What happens when something goes wrong. The end product. The user experience. The craft parts. For the first time in my life as a tool builder, the code was not the bottleneck. The design was.

What I bring, what the AI brings

I want to be clear about what this kind of work requires, because it is not "prompt the AI and collect the plugin."

I'm not a software engineer. I have a foundational grasp of Python, a bit of Modo scripting, and enough HTML to be dangerous. Before Claude Code, the plugins I built were small and brittle and took weeks.

What I do bring is harder to teach. 12+ years of professional 3D production experience. A pretty good sense of what a workflow should feel like. I can tell when a tool is asking for too many clicks, when a UI is getting in the way, when the friction is hiding in a place most engineers would never think to look. And I can say clearly what I want, and recognize when what I got back isn't it.

Those are the things that make AI-assisted development actually work. Not the typing of the code. The knowing what to build, and the knowing when the code is wrong.

What's actually changed

The old ceiling on plugin development was technical. If you couldn't write it, you couldn't have it. The new ceiling is shaped differently. It rewards people who have spent years inside a craft. People who notice the small moments of friction in a day that a newcomer wouldn't even see. People with the slow-accumulated taste that only comes from thousands of hours doing the work.

In the right hands this is a big deal. Experienced artists are quietly building internal tools that used to require a dedicated software team. I've watched designers assemble plugin suites in a few weeks that streamline the exact pain points they've been living with for a decade. None of them are engineers. All of them know their craft cold.

What comes next

I don't think "AI replaces artists" is the interesting story. It's too easy, and it's not what I'm seeing on the ground.

The real shift, the one I find exciting, is that the tools we use are about to become personal. Not generic. Not one-size-fits-all. Built by the people who use them, shaped around how those people actually work, updated when the work changes. A new kind of craftsmanship is showing up, one where deep domain experience and AI-assisted development combine into something neither could produce alone.

If you're an experienced artist wondering whether any of this matters for you: it does. The tools you've wanted for years are probably within reach now. The investment is smaller than you think. And the first plugin you build to save yourself ten minutes a day will pay for every hour you spent learning.