Postyourbuildideas

Developer Tools

A public board for the app ideas you'll never get around to building. Post one and someone with a free weekend picks it up — or take one if you're staring at an empty editor. Every idea comes with a copy-ready build prompt you paste straight into Claude Code or Cursor. No account, no signup, no claim on what gets built. Free for everyone. Funded by weekly sponsor slots from $5 to $100, with every click count published so a sponsor can check whether it worked. Six days old. 34 ideas on it so far.

2
Roasts
6.5/10
Avg. score
50%
Would pay

What the community said2

Steven SecretiSteven Secreti
Helpful7/10Wouldn't pay

What works

More than the homepage lets on, honestly. I almost roasted you for having no prompts at all, then clicked into a detail page and found a real generated build prompt: idea-specific context, an MVP scope that says "smallest version, nothing more," explicit empty/error/offline state handling, even a pricing model to build toward. That's a genuinely useful artifact hiding behind a click. The zero-friction bet (no account, no fee, no claim) is the right one, and 40 prompts copied in 6 days says people want these.

The biggest problem

Your best asset is invisible. The list view is 2-line pitches, so a skeptical visitor bounces off "a photo of the fridge, and dinner is decided" without ever learning the prompts exist; I nearly did. And the loop still isn't closed: 40 copied, 0 shipped, and no way for a builder to come back and attach what they built. I hand prompts to coding agents daily and I think the template has a ceiling too, since all three I read share the same skeleton with no target user and no done-criteria beyond "deployable in a day," which is a stack constraint, not a definition of done. Also the "automatic tug device" sitting unvetted next to real ideas quietly tells serious builders this is a joke board.

Top suggestion

Surface the prompt before the click. Put the first couple lines on the card, or a "view build prompt" teaser, so the list view sells the thing that makes you different from a wish list. Then give copiers a way to attach a deployed link to the idea they built from; one complete chain, idea to prompt to live URL, is worth more than the next 40 posts, and it's the only thing that moves that 0.

Sign in to join the discussion.

Martin SailliezMartin Sailliez
Helpful6/10Would pay

What works

The roast form is the real asset. Forcing "biggest problem," "would you pay," and a 1-10 score turns vague niceness into feedback a founder can act on. I've written a dozen of these today and the structure does the heavy lifting, that's a genuinely good product decision, not a comment box.

The biggest problem

Your leaderboard rewards volume, and volume is the enemy of the honesty you're built on. "Top roasters" ranks by raw count (12 roasts, 3, 2...), not by whether any of them helped. So the incentive it quietly creates is: dash off ten shallow "nice work, 7/10" roasts to climb, instead of one that actually lands. Your own thesis is that roasting-for-the-prize kills roasting-for-honesty, but the ranking IS a prize, and right now it pays out for quantity.

Top suggestion

You already collect "helpful" marks. Make THAT the leaderboard. Rank roasters by helpful votes received, not roasts given. Same gamification, but now it rewards depth and punishes drive-by scores, without adding any reward you're philosophically against.

Sign in to join the discussion.

Want to roast back, or ship your own?

Reading is open to everyone. Posting a roast or a project takes a free account.

Sign in, it's free