Can LLM Act as Static Analyzers? An Empirical Study on Django Backend Code

This repository contains the anonymized replication package for the study submitted to MSR 2026.It includes: The validated dataset of code-quality issues (70 confirmed issues) Raw outputs from static analysis tools (SonarQube, DeepSource, Prospector) Raw LLM outputs (anonymized) The complete prompt template and code-chunking procedure used in the study A README file with instructions to reproduce the evaluation No author-identifying information is included in this package.All file paths and metadata have been anonymized to comply with the MSR open-science policy.