CWE
A standardized enumeration system of software weaknesses maintained by MITRE. It assigns a unique identifier to each security weakness and is widely used for vulnerability management and integration with security tools.
Timeline (last 90 days)
The case study's method: first run deterministic scanners (Gitleaks, Trivy, Semgrep, Hadolint) to generate a ranked findings document, then layer on three model stages—the discovery stage treats scanner output as clues and reads source code against the OWASP and CWE frameworks, the chain investigation stage combines individual findings into attack paths, and the triage stage confirms whether the findings are viable.
The LSCBench benchmark contains 450 tasks, derived from 150 real vulnerabilities, covering seven languages and 24 CWE categories.
CWE is maintained by MITRE and is a standardized enumeration of software weaknesses, each with a unique number, such as CWE-89 for SQL injection and CWE-79 for XSS.
All Facts (3)
The case study's method: first run deterministic scanners (Gitleaks, Trivy, Semgrep, Hadolint) to generate a ranked findings document, then layer on three model stages—the discovery stage treats scanner output as clues and reads source code against the OWASP and CWE frameworks, the chain investigation stage combines individual findings into attack paths, and the triage stage confirms whether the findings are viable.
50%UnverifiedThe LSCBench benchmark contains 450 tasks, derived from 150 real vulnerabilities, covering seven languages and 24 CWE categories.
50%UnverifiedCWE is maintained by MITRE and is a standardized enumeration of software weaknesses, each with a unique number, such as CWE-89 for SQL injection and CWE-79 for XSS.
50%