Static and Runtime Analysis of Auto-Generated Application Logic in Low-Code Systems

Abstract

Low-code platforms use auto-generated logic to speed up enterprise software development, but this can create hidden defects, inconsistent rule execution, and weak validation. Existing approaches focus mainly on efficiency and provide limited support for verifying generated logic in complex environments. This study proposes a unified verification framework combining static analysis and runtime validation to improve correctness and reliability in low-code systems. The framework formalizes generated workflows, applies constraint-based static checks to detect structural inconsistencies, and uses runtime monitoring to identify execution anomalies and context-dependent deviations. Evaluation across different logic complexity levels shows improved defect detection accuracy, stronger rule coverage, and acceptable validation latency compared with standalone verification methods. The findings show that static and runtime techniques complement each other by addressing both pre-execution and dynamic validation challenges. The proposed framework supports more reliable, compliant, and correct enterprise low-code applications.