Qexploit

<p dir="ltr">QExploit: Characterizing Quantum Advantage for Quantitative Software Exploitability Analysis</p><h2 dir="ltr">Quick Start</h2><pre><pre># Install dependencies<br>pip install -r environment/requirements.txt<br><br># Run complete test suite<br>python tests/test_all_components.py<br><br># Reproduce full experimental study<br>python experiments/full/run_full_study.py<br><br># Generate publication tables and figures<br>python src/analysis/generate_tables_figures.py<br></pre></pre><p>=======</p><p dir="ltr">QExploit: Bounded Predicate Counting Feasibility Study</p><p dir="ltr">QExploit is an offline research artifact for studying projected model counting and conditional quantum-resource estimates on <b>manually constructed bounded predicates</b> linked to cases in the public SB-Heists artifact.</p><h2 dir="ltr">Scope and evidence status</h2><ul><li>The 32 predicates contain 16--24 attacker-controlled bits.

They are low-dimensional abstractions, not formulas extracted from Solidity/EVM execution semantics.</li><li>Exact counts and ApproxMC results characterize only the declared predicates and domains. They do not estimate real-world exploitation likelihood.</li><li>All eight unchecked-call records share one Boolean predicate, so the nominal 32 rows are not 32 independent semantic observations.</li><li>Quantum results are analytic query and category-template resource estimates.