What problem does it solve? Regulatory prior-art surveys can contain fabricated citations, misclassified instruments, or quietly resolved conflicts that look plausible but are wrong. This Skill acts as the judgment half of a two-part quality gate, reviewing survey artifacts against a fixed set of numbered conditions so defects are caught before an architecture or compliance decision is built on them. ## Core Features & Use Cases - Condition-based review: Judges artifacts against 35 numbered conditions (C1-C35) covering scope maps, search outputs, extract records, and the regulatory register, defined in references/conditions.md. - Fabricated citation detection: Prioritizes catching obligations the instrument text does not carry, unresolved identifiers, and text_retrievable states the fetch never reached. - Structured verdict output: Emits findings each naming their condition, followed by exactly one terminal VERDICT line of approve or revise, with UPSTREAM labels for defects outside the artifact author's control. - Use Case: After a regulatory-prior-art-survey run produces a scope map and search output for a telehealth product, run this reviewer to verify per-angle verdicts, verbatim query recording, and claim-versus-quote boundaries before approving the survey. ## Quick Start Use reviewing-regulatory-prior-art-survey to review the search output artifact from my regulatory survey and give me the verdict with findings.
