This study designs and validates AISA–L, a four-layer Agentic AI Strategy Architecture (Perception, Cognition, Strategy, Action) that converts a previously validated comprehensive portfolio of 110 airline logistics Key Performance Indicators (KPIs) [1] into autonomous, auditable, sustainability-aligned decisions. The novelty lies not in enumerating KPIs, but in their real-time agentic orchestration within a closed governance–optimization loop. Addressing the persistent gap between descriptive dashboards and adaptive execution, the research operationalizes KPI governance (threshold analytics, anomaly detection, bias auditing, explainability), multi-objective optimization (cost, resilience, carbon intensity, inventory balance), and disruption response (AOG rerouting, maintenance reprioritization). A mixed-methods design science approach integrates purposive expert elicitation with digital twin simulation contrasting a baseline manual governance model against the agentic configuration. Empirical results show a 22% improvement in forecast accuracy, ≈11% reduction in turnaround time, 1.6 percentage point increase in aircraft dispatch reliability, 4.8% CASK reduction, ≈9% inventory turnover uplift, 18% faster disruption recovery, 6.3% decline in CO₂/RTK, and a 7-percentage point rise in sustainable aviation fuel utilization, alongside zero material bias incidents and enhanced data timeliness. Theoretically, the study reframes KPIs from evaluative endpoints to real-time control variables within a cyber-physical logistics governance loop, extending digital maturity and ethical AI discourse. Practically, it delivers an implementable blueprint for Chief Logistics Officers and regulators to embed sustainability, resilience, and ethical compliance into continuous optimization. Recommendations include phased agent deployment, constraint-based ESG integration, lineage-centric data governance, and capability KPIs for human–AI co-leadership.