Get Started
Home
Topics
Search
Library
Research questionHow can vulnerability-fixing history be reused to detect and repair similar undisclosed flaws in other code?Vulnerability databases and fixing commits explain how known flaws were removed, but their code-level lessons are difficult to apply systematically to code without a known advisory. The challenge is identifying genuinely similar unsafe conditions and validating that proposed repairs address them without introducing new behavior changes.
AI
AI Agents
Code Generation & Program Synthesis
Evaluation & Benchmarks
Technology
Latest papersRecent research connected to this question, newest first.The History Is the Detector: Executing CVE Patch History, End-to-EndThe source describes BUGSTONE-E2E, which mines verified CVE fixing commits into rules organized by CWE and programming language, then applies lightweight analysis, LLM-based inspection, runtime verification, and differential testing across 14 programs. Its reported evidence covers 19,325 high-severity CVEs from 2022–2026, yielding 1,033 rules and runtime evidence for 644 findings; the applicability of those results beyond the studied programs and extracted rule families is not established.research paper · Sep 4, 2026
Related questions
How can code-repair models fix bugs without unnecessarily rewriting surrounding code?How can we detect when LLM-decompiled code diverges from original behavior or erases disclosed vulnerabilities?How can coding agents report defective test infrastructure instead of exploiting it to pass?How can retrieval-augmented code generation resist targeted vulnerabilities from one task-matched poisoned artifact?