analysis_root is the experimental output. Python_CWE_Bench is the vulnerability database source file and its compiled database. Security-aug is the enhanced SQL file. Experiment - GenerateQL.py is used to call CQLLM to generate SQL code and save it to a SQL file. Experiment - ExecuteQL.py is used to determine whether the SQL file generated by CQLLM can be successfully executed. Release includes the CWEFix database and query log files.