Developed a lightweight validation feature for the xupefei/spark repository, focusing on enhancing the Spark SQL optimization pipeline. The solution introduced a mechanism to check if a logical plan becomes unresolved after each optimizer rule, enabling early detection of potential bugs during query optimization. By surfacing unresolved plans immediately, the approach improved debugging feedback and made it easier to triage optimization-related issues. The validation was carefully designed to be non-intrusive, ensuring no impact on runtime performance. This work leveraged Scala and Spark, demonstrating backend development skills and contributing to the overall robustness and reliability of the optimization workflow in Spark.