Most idea validators are feel-good machines: paste an idea, get a "promising" score, then build for months on a vibe.

WhittleOS is the opposite — deliberately kill-biased. Ask it about an idea you already have and it returns build / test-first / drop, with the reasons. Or let Discovery source real candidates for you across a market and rank them against your founder profile.

What makes the verdict checkable rather than flattering:

• It links the problems it cites to the page each was read on, or labels them as its own estimate. The published run documents 56 problems, 32 of them with a link you can open.

• The 0–100 score is computed in code from a fixed weighted rubric across seven categories, not asked of a model.

• When a market is genuinely weak it hands back nothing rather than a score. That published run took 69 candidates, killed 36 on deal-breakers, and its best surviving idea scored 52 — "test it first", not "build it".

Validation is no-call and async by design: a landing page and a real CTA, never "go book 20 discovery calls".

Seventeen analysis modes in total, with Discovery as the flagship. Prepaid credits, pay per decision, no subscription — two free credits on signup, three calculators at whittleos.com/tools with no signup at all, and the full published run readable at whittleos.com/sample-discovery.