wrong-altitudelisted

# An applied-layer answer hides the governing result ## Problem A search that fails prompts a wider search. **A search that succeeds shallowly does not.** This is the whole mechanism. When a prior-art sweep returns material that is relevant, correct and citable, every quality signal reads green: sources found, claims supported, citations recorded. The sweep is *done*. Nothing in the process asks whether the answer sits at the right depth, because nothing has gone wrong. So the sweep terminates one layer above the result that actually governs the problem — typically in applied papers, vendor documentation, practitioner writing or case studies, all of which are instances of a theory nobody went looking for. The cost is asymmetric and delayed. A wrong answer gets caught. A *shallow* answer gets built on, cited forward, and hardens into the frame everything downstream assumes. ## Context / Trigger Conditions Fire this **after a sweep succeeds**, not after one fails. - A prior-art or literature sweep returned usable sources and you are about to act on them. - The sources are applied: schema docs, evaluation frameworks, tooling papers, practitioner blogs, vendor documentation, case studies, "how we did X" writeups. - Two or more sweeps on the same question both returned the same layer. **Agreement between shallow sweeps is not corroboration — it is the same blind spot run twice.** - You are about to write that something "has no prior art", "does not reduce to existing w