Property-based testing tools such as Hypothesis can turn a formal grammar into a generator of syntactically valid inputs, but the generator itself is typically static: hand-tuned once and reused unchanged for the rest of a fuzzing campaign. We study whether a large language model (LLM), given only a formal grammar and parser-level feedback – with no coverage instrumentation – can iteratively refine such a generator. We build a fully black-box pipeline targeting the cJSON parser (pinned at v1.7.19): a baseline Hypothesis strategy derived directly from an ANTLR JSON grammar generates inputs, a sanitizer-instrumented harness classifies each as accepted, rejected, or crashing, and a refinement loop feeds an LLM a summary of three observable proxy signals – acceptance rate, the structural diversity of the shapes produced, and the distinct rejection signatures seen – asking it to propose a revised generator. Every proposal is statically sandboxed (AST-checked, import-restricted) and validated before it ever touches the target binary. Across fifteen independent, seeded runs of up to five refinement iterations each, LLM-guided refinement raises mean acceptance rate from 58.2% to 97.1% relative to a static baseline, using no coverage data at all. Structural fingerprints and rejection signatures are reported as descriptive signals; because the refined arm executes more iterations, their aggregate counts are not treated as like-for-like effects. A follow-up ablation over which subset of this feedback the proposer sees – counts alone, counts with rejection signatures, or the full signal – finds acceptance-rate gains are broadly similar across the three conditions, with full feedback preserving more rejection-signature diversity than counts alone (four usable runs per condition; reported descriptively, not tested). We additionally run the identical, unmodified pipeline against a second parser (parson) for five real iterations (3,000 sanitizer-instrumented executions); no crash was found, and we report this as a negative result, supported by manual review of the highest-risk code paths, rather than claim a bug that was not observed. We discuss how grammar-implementation mismatches – for instance, duplicate JSON object keys accepted by one parser and rejected by the other under an identical grammar – are themselves a source of fuzzing-relevant signal that a purely formal grammar cannot supply, and we release the full pipeline, harnesses, and reproducibility tooling.
Ziyad Mohammad Mansy Ibrahim· Zenodo (CERN European Organi...· 0 citations
Property-based testing tools such as Hypothesis can turn a formal grammar into a generator of syntactically valid inputs, but the generator itself is typically static: hand-tuned once and reused unchanged for the rest of a fuzzing campaign. We study whether a large language model (LLM), given only a formal grammar and parser-level feedback – with no coverage instrumentation – can iteratively refine such a generator. We build a fully black-box pipeline targeting the cJSON parser (pinned at v1.7.19): a baseline Hypothesis strategy derived directly from an ANTLR JSON grammar generates inputs, a sanitizer-instrumented harness classifies each as accepted, rejected, or crashing, and a refinement loop feeds an LLM a summary of three observable proxy signals – acceptance rate, the structural diversity of the shapes produced, and the distinct rejection signatures seen – asking it to propose a revised generator. Every proposal is statically sandboxed (AST-checked, import-restricted) and validated before it ever touches the target binary. Across fifteen independent, seeded runs of up to five refinement iterations each, LLM-guided refinement raises mean acceptance rate from 58.2% to 97.1% relative to a static baseline, using no coverage data at all. Structural fingerprints and rejection signatures are reported as descriptive signals; because the refined arm executes more iterations, their aggregate counts are not treated as like-for-like effects. A follow-up ablation over which subset of this feedback the proposer sees – counts alone, counts with rejection signatures, or the full signal – finds acceptance-rate gains are broadly similar across the three conditions, with full feedback preserving more rejection-signature diversity than counts alone (four usable runs per condition; reported descriptively, not tested). We additionally run the identical, unmodified pipeline against a second parser (parson) for five real iterations (3,000 sanitizer-instrumented executions); no crash was found, and we report this as a negative result, supported by manual review of the highest-risk code paths, rather than claim a bug that was not observed. We discuss how grammar-implementation mismatches – for instance, duplicate JSON object keys accepted by one parser and rejected by the other under an identical grammar – are themselves a source of fuzzing-relevant signal that a purely formal grammar cannot supply, and we release the full pipeline, harnesses, and reproducibility tooling.
Ziyad Mohammad Mansy Ibrahim· Zenodo (CERN European Organi...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
