Analyze agent sessions against verifier checklists, detect friction points, and create structured verifiers from skills and docs. Produces per-session verdicts and aggregated quality reports.