3 minutes reading time
So Roo Code got taken out back and shot today.
The VS Code extension, the little local agent that three million people installed. Dead. They are archiving the repo on May 15. The reason is exactly what you probably suspect. Giving away local software doesn't pay back venture capital.
They are pivoting to a thing called Roomote. It lives in the cloud and hooks into Slack. Because selling an agent platform to entire development teams is where the real money sits. The funny part is that the prototype for Roomote was literally just them running Roo Code in cloud containers to speed up their own pull requests. Now they want to dress that process up, slap some enterprise compliance labels on it, and sell it to CIOs. You can't squeeze predictable monthly revenue out of a local desktop tool. So they just kill it.
You might think they'd hand the project over to a foundation. Let the community maintain it. Nope. They absolutely refuse.
Sure, the code is open source. Anyone can technically fork it and maintain it under some random new name. But why not just transfer the actual project—the brand, the repo, the momentum—to a foundation? I didn't initially know.
But then I read the developer threads. Roo actually started as a fork of Cline because Cline wouldn't merge an auto-approve PR. The founders spun it off, added features, and now they claim it is entirely "theirs." Giving the brand to a foundation apparently ruins their startup origin story. They genuinely care more about controlling the mythological narrative of being the "creators" than letting the tool survive under its established name. I guess it's easier to just lock the doors and walk away. A core developer on Reddit actually tried to argue this wasn't a cash grab. Right. Sure thing, man. Whatever helps you sleep at night.
I shouldn't even be annoyed. This is just the standard playbook now. Build a free toy. Bait three million users into your ecosystem. Figure out your margins are basically zero because local agents don't generate cash. Pivot to selling bloated enterprise subscriptions.
If this trend continues, the whole ecosystem is going to collapse. Companies extract value from open source, build a tool, automate everything, and put zero structural support back in. Then the VCs demand a return on investment and they pull the plug on the free ride.
Go use Cline, or Kilo, or whatever local project hasn't taken Series A funding yet. Enjoy it while it lasts. Because right now, "open-source AI tooling" is just a disguised waitlist for a closed enterprise product.
It keeps happening. And it's thoroughly exhausting.