Bump - The starter kit that ships your go-to-market too

by
Bump is a production starter kit for founders who build with AI. One codebase ships a consumer app, customer workspace and admin on iOS, Android and web (Expo), plus a marketing site on the same design tokens. Supabase, Stripe + in-app billing, streaming AI chat and voice. The difference: go-to-market engines in the box. Inclusive of cold outreach, a Reddit copilot, AI quote decks, and an agent pack so Claude Code or Codex sets it up with /setup. $299 once, unlimited projects.

Add a comment

Replies

Best
Maker
📌
Hey Product Hunt 👋 I'm Justin, a solo founder in Australia. I spent the past year building Koora, a care-sector SaaS. Somewhere along the way, the engine underneath became the thing I was just as proud of: the auth, the payments, the admin, and especially the go-to-market machinery wired into our own admin app. So I forked it. A couple weeks of stripping out everything care-sector turned ten months of production code into Bump: your product on web, iOS and Android from one Expo codebase, a marketing site sharing the same design tokens, Supabase with RLS on all 58 tables, Stripe plus in-app purchases, and the part no starter kit ships: growth engines. Cold outreach as a Monday-morning approval queue, a Reddit copilot that drafts replies in your voice, quote decks a prospect accepts (with Stripe integration) in one click. AI drafts, you approve, nothing sends without a human. It's built to be worked on by agents. Open the repo in Claude Code or Codex, type /setup, and your agent does the setup itself. And I won't pitch harder than this: the site has a prompt you paste into your own AI to work out whether Bump is worth it for your project, including the case against buying. $299 once (launch price), unlimited projects. Every sale funds Koora staying bootstrapped. I'll be around all day, so ask me anything about the stack or the growth engines. Feedback very welcome.

Spent an evening poking around the starter and the design token reuse across web and mobile actually holds up, which is rare. The Reddit copilot feels useful without being cringey too.

 "Useful without being cringey" might be the nicest thing anyone's said about the Reddit copilot. That was the whole design constraint: it drafts, you approve, and it's tuned to add to the conversation rather than drop links. Really good to hear the token reuse held up under an actual poke-around too. Simple idea, but it does a lot of work.

The design tokens approach is smart, saved me hours fiddling with the marketing site to match the app. Setup with /setup was straightforward too.

 Thanks Kamil, good to hear it's working for you too. Sharing tokens between the Next.js site and the Expo codebase is such a simple idea but it saves so much fiddling. Glad /setup behaved itself as well.

Spent an afternoon with the kit and the Supabase + Stripe wiring was already preconfigured, which saved me a whole weekend. The /setup command actually worked with Claude Code and had the auth flow running in minutes.

 Great to hear, Salim. Auth running in minutes is exactly the outcome we were aiming for, and it's good to know /setup held up on someone else's machine, not just ours. If you hit anything rough as you go deeper, I genuinely want to hear about it.

One thing I'd love to see is a built-in feature flag system tied to Stripe plans, so I could gate premium AI features per tier without wiring up a separate tool. Would save a ton of time during the first launch.

 That's a really good idea, Sevda. The Stripe + IAP billing is already wired end to end, so the entitlement data is there, and a feature-flag layer on top of the plan tiers is the natural next step. Adding it to the list. The nice thing about a kit though is you don't have to wait for me: the agent pack knows the billing code, so asking Claude Code to gate a feature per tier is a small job, not a separate tool to wire up. Thanks for taking the time to write it up.

The shared design tokens across web and mobile saved me a ton of setup time. Curious how the cold outreach integration holds up against dedicated tools like Instantly.

 Thanks Meral. Good to hear you're liking the shared tokens too. It's something super simple but really powerful.

 On outreach: it ships wired into Lemlist, with the part dedicated tools don't do. AI drafts the outreach from your product's actual context, you approve, and nothing sends without a human. But the real answer is that you own the code. If you prefer Instantly, point Claude Code or Cursor at the outreach engine and swap the sending layer. It's a clean integration, not a black box. That's the whole benefit of a kit like this over a SaaS: you're never locked into our tool choices.