Experimental benchmark demonstrates operational provenance tracking under identifier failure across Kubernetes environments, indicating bounded feasibility for cross-plane auditing.
Cloud-native operations scatter audit-relevant evidence across delivery and runtime systems. This preprint presents EACP 1.3, a domain-specific operational-provenance profile and materialized retrieval index that retains native evidence pointers, uses typed and scoped exact links, and abstains when the selected identifier is missing or ambiguous. A 30-seed adversarial campaign measures degradation under identifier failure, while paired SQLite ablations isolate retrieval, storage, and ingestion costs. A controlled GitHub Actions-to-Kubernetes study preserves an initial 0/9 cohort that exposed a lifecycle-invalid assertion and separately reports 9/9 confirmatory first-attempt runs across Kubernetes v1.34.8, v1.35.5, and v1.36.1 after a prospectively committed narrow amendment. The results support bounded feasibility and controlled compatibility—not causal inference, source truth, production reliability, independent replication, or distributed-scale performance. Preprint; not peer reviewed.
No takes yet. Share an insight, caveat, or question.
Obede Bessa Rocha da Silva (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: