I’m Max.

I built this because I have a "Yes Man" problem. Whenever I validate ideas with friends or standard LLMs, they are too polite. They focus on why an idea might work, rather than why it will fail.

I tried prompting ChatGPT to be mean, but it often drifts back into being helpful.

So I built Crucible. It orchestrates several specialized agents (a CFO focused on unit economics, a CTO, a Innovation Manager) that debate your idea simultaneously.

It’s not just a chat; I spent a lot of time tuning the system prompts to force them to look for logic gaps, market saturation, and risks rather than just summarizing the pitch.

It costs $5 (one-time). I hate monthly subscriptions, so I just made it a single payment.

I’d love to know if the feedback feels harsh enough, or if I need to dial up the temperature. Roast my work that’s the whole point.