The contents of this repository are broken into three directories:

• The tool implementation
• Papers
• Results from paper

To re-run the experiments that were presented in the paper or to run your own, view the README in the ErrorSpecifications directory. Please note that running the interleaved analysis with the LLM is inherently non-deterministic, so the per-run results will likely be different.

The paper Interleaving Static Analysis and LLM Prompting that was published at SOAP 2024 is included at paper/soap24_paper.pdf. The paper Interleaving Static Analayis and LLM Prompting with Applications to Error Specification Inference from STTT 2025 is included at paper/sttt25_paper.pdf.

We have included the raw experimental results for the inferred error specifications in the results directory. As the interleaved analysis is non-deterministic, we include these results separately as re-running the analysis may yield different results.