Assessment method and scorecard for AI agent deployments
An impact assessment method and scorecard at the deployment and agent layer, using points of relation and points of control (PoR/PoC) as the working unit of analysis: 56 metrics in 10 domains, maturity levels capped by evidence grade, release gates that fail closed, and no aggregate score.
Not validation of the metrics or thresholds; not yet a benchmark or certification.