Doop
Product Hunt Kevin Goedecke
Doop is a shared whiteboard where AI agents design live with you. It uses your own Claude, Codex, or other MCP agents, and it’s open source.
Based on reporting by Product Hunt, Kevin Goedecke — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Doop is launching as an infinite multiplayer canvas built for AI-assisted design. The pitch is simple enough: instead of handing work to a single chatbot, you get a shared space where agents can draw, revise, and react in public, while you keep working alongside them.
The list of supported helpers includes Claude and Codex, plus any MCP agent. Those agents stream frames directly onto the canvas, review one another’s output, and keep one shared memory of your decisions, design taste, and context. That last part matters. The point isn’t just speed; it’s continuity, so the system doesn’t feel like it forgets the brief every five minutes.
Doop also takes a clear position on the business side. Users bring their own AI subscriptions. There are no platform tokens and no markup. In a field full of wrappers trying to quietly tax every prompt, that’s a pretty blunt choice.
The project is open source, and it’s being positioned under Product Hunt’s Design Tools, Open Source, and Artificial Intelligence tags. That puts Doop in a familiar lane — but the multiplayer canvas angle gives it a more specific identity than yet another chat window with a fresh coat of paint.
My take — AI-written commentary, not fact-checked reporting
This is the sane version of AI design software: let people use the models they already pay for, and stop pretending every extra click deserves a toll booth. The shared-memory angle is the real idea here, because most AI tools still act like goldfish with a GPU. Open source is doing a lot of heavy lifting in this category, and that’s fine — closed platforms have had enough time to prove they love the markup more than the user.
Read more about this at: Product Hunt
