VibeCheck.exe is an AI thinking partner that stress-tests a startup idea before you spend a weekend building it. Instead of agreeing with you, it asks the questions a sharp cofounder would ask, then tells you to sharpen the idea, redirect it, or drop it. It's built by Lex Hamilton at charmthirteen and is currently in preview.
Why does idea validation matter more than build speed now?
Vibe coding compressed the time between an idea and a working prototype down to an afternoon. That changed what's actually scarce. Building was the bottleneck for years, and now judgment is. Most founders and vibe coders who ship something nobody wants didn't build it badly. They built the wrong thing, quickly, because nothing in their process asked them to stop and check first.
General AI assistants don't fill that gap. An assistant tuned to keep you engaged will find something to like about almost any idea, because agreement keeps the conversation going. One-time validation reports have the opposite problem, since they hand you a verdict and forget you exist the moment you close the tab. Neither one functions like a partner who knows your context and will tell you something you don't want to hear.
What is VibeCheck.exe?
VibeCheck is that partner. You describe an idea in plain language, and it pushes back with the questions that determine whether the idea survives contact with a real market. Who is this for, and what problem does it actually solve for them? Does something like this already exist, and how crowded is that space? Are you close enough to the problem to see what a casual observer would miss?
VibeCheck then gives a direct answer, which is to sharpen the idea, redirect it toward something more promising, or let it go. The redirect is the point. A tool that only ever says yes is a mirror, and founders already have plenty of those.
How does VibeCheck decide if an idea is worth building?
Before VibeCheck responds to any idea, it checks what's already out there, so its feedback is grounded in the current market rather than a confident guess. That single step is what separates a strategic partner from a flattering one, because the pushback is based on evidence, not vibes.
What is the AI Collaboration Pyramid?
VibeCheck runs on a framework Lex teaches called the AI Collaboration Pyramid, built on three levels that stack on top of each other. (1) Ask with curiosity, drawing out the why behind the idea before rushing toward a verdict. (2) Explore with intention, iterating on the idea, probing its weak points, and pulling in what the market is already doing. (3) Refine with judgment, where the human stays the decision-maker; VibeCheck informs the call, it doesn't make it.
VibeCheck is that framework built into something usable, wrapped in a Y2K desktop window because working through a hard decision doesn't have to feel like homework.
Should you build your idea? Three questions to ask first
Before opening any tool, run your own quick check with these three questions. Who specifically is this for, and have you actually talked to them? Does something like this already exist, and why would someone pick yours over it? Are you close enough to the problem to notice what everyone else is missing?
If you can't answer all three with specifics, the idea isn't ready to build yet. That gap between "I have an idea" and "I can answer these three questions" is exactly what VibeCheck is built to close.
Is VibeCheck available yet?
Not yet. VibeCheck is in preview. A phone version exists and is currently Lex's alone while she tests it, and a web version is coming under charmthirteen. Both versions will share the same underlying logic even as they live in different places.
Sharing an unfinished product on purpose is the whole thesis behind VibeCheck, which is to validate the interest before you finish the thing. So the question VibeCheck would ask about any idea gets asked here too. Would you actually use this?