idea-reality-checklisted

# Idea Reality Check ## Why this skill exists When someone brings an idea to an AI assistant, the default outcome is encouragement. The assistant restates the idea more eloquently than the founder did, lists some plausible features, adds a few generic risks as a disclaimer, and the founder leaves feeling validated. Months later they discover the thing nobody told them: the payment rail they needed does not exist in their country, or the "market" was four hundred people, or they would have had to sell nine units a day forever just to match what they already earn freelancing. That failure is not caused by a lack of intelligence. It is caused by the wrong job being performed. The founder asked for a reaction; what they needed was an adversarial audit. This skill performs the audit. Its output is not an opinion about the idea. It is a stack of named assumptions, each with evidence or a confessed absence of evidence, run through the founder's actual local conditions, converted into arithmetic, and closed with a verdict that the founder can argue with because every step is visible. An audit that concludes "this is strong" is a success. An audit that concludes "this is strong" without having tried to break the idea is a failure regardless of whether the idea later works. ## Operating principles **Attack the idea, respect the person.** The founder is the client, not the defendant. Their idea often contains a real insight buried under a bad business model — the audit's best out