Propane gives your product team and agents one connected, always-current view of your customers. Automatically collected from all your tools. Collaborate on a shared canvas. Commit straight to any coding or design agent. Secure, maintained, always on. You just build products people love.
Nice work, excited to try it out 🙌 What was the biggest aha-moment when building Propane?
@stockmarr Yes please do! That teams don't need to mange MD files, context, skills and all teams are work from the context, that is automatic collected. They keep telling us they spend so many token, time to mange all of this. With propane that is now simple!
@stockmarr 🙏🙏🙏
Congrats on the launch. that website is super cool - what tools did u build it w. ?
@rayhan_khilji thanks Rayhan!! that question will have to go to our webmaster general @atherkildsen 😎 I also agree it's super cool! did you see the interactive product hunt kitten? kinda stealing the show.. 🐈;)
@rayhan_khilji @atherkildsen used @Figma @Cursor @Vercel and a big bucket of tokens 😂
Product's Cursor moment, love it. The shared canvas is the part I'd actually use
@yanshuo appreciate the support Mogu!!
For sure, canvas is free to try, would love to hear your thoughts! don't forget you get 3 months subscription free with the above code!!
@yanshuoThat would be amazing to get that level!
@yanshuo love that website! We’re here if you need a guided tour 🙋♂️
The day has finally come 🔥
Super excited to get this out and see what problems you'll solve with it.
Being on a team where everyone believes so deeply in what we're building tells me this was made with real heart - and I can't wait for you to experience it. Go check out what we made for you.
@nicolai_harms very fun to tag-team building some core stuff on this one together!🚀🚀🚀
@nicolai_harms @benfleming Get a room 😆
@nicolai_harms @atherkildsen 😳😳
Great idea. what part of the product do users usually understand the fastest?
@rose_florean the Canvas is very compare some experiences they have tried in notion, claude etc. So that is fastest from our data.
I'm curious how the agent side handles freshness. When we pipe customer context into coding agents, the thing that bites us is the agent acting on a complaint that was already resolved, or treating a three-month-old signal as current. Does Propane scope what slice of the customer view an agent sees per task, or does the whole always-current view land in the context window?
Customer context feels like one of the biggest bottlenecks for useful AI agents.
Curious—does Propane end up replacing multiple context sources, or does it work best as the layer connecting everything together?