review-adversarial

Adversarial peer-review systematically stress-tests document claims by applying multiple hostile-but-fair reviewer personas. This approach targets argumentative flaws, logical fallacies, and evidentiary gaps to identify manuscript vulnerabilities before publication.

Yes, you can audit repository file globs. The process resolves scope, enforces guards for empty scopes, and sequentially runs persona audits to stress-test claims, logic, and clinical relevance across matched files, aggregating per-persona findings.

The vulnerability assessment report aggregates per-persona findings and detects cross-persona convergence. It highlights cross-persona vulnerabilities and resilient claims, outputting the consolidated markdown report to your XDG_RUNTIME_DIR.

Yes, clinical relevance testing is a core feature. One of the six ordered adversarial personas specifically targets clinical recommendations to ensure research chapters withstand rigorous stress-testing and maintain medical relevance.