Hey PH 👋

well, what can I say.

Working with AI in early 2026 was a real pain in the neck. I was bouncing between Claude, Codex and Cursor all day, and every time I switched tool or ran out of tokens I had to explain the same goal, the same plan and the same decisions all over again. One day I woke up tired of repeating myself a thousand times, so I started building Goodboy.

At first it was just a shared context between agents: one local place that holds what you’re working on and hands it to whichever agent runs next. Then every time something annoyed me in my daily workflow, I turned it into a feature. That’s how it grew into a real ADE. Agents run in parallel on isolated worktrees, workflows send each step to the right model, and it’s integrated with the tools I use every day, like GitHub and Linear, so issues and PRs sit right next to the agents working on them.

A few things I care about: it drives the official CLIs on the subscriptions you already pay for, or on your own API keys if you prefer. Everything stays on your machine. And it’s built in public.

I’m building it solo next to my day job, so feedback really matters. What would help most:

where the onboarding gets confusing
which integration you’d want next
what breaks, because something will 🐶

And if you like the project, a ⭐️ on GitHub would mean a lot

Thanks for checking it out, I’m curious to know what you think :)

Oh, now you can try a live demo from your browser XD