vulnerability-pattern-matcher

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill contains no malicious code, obfuscation, or dangerous commands. It functions as a guideline for security auditing.
  • [NO_CODE]: The skill consists entirely of markdown instructions and metadata, with no associated executable code or scripts.
  • [EXTERNAL_DOWNLOADS]: The skill references an official GitHub repository from the Coalition for Secure AI (COSAI-Oasis) for documentation purposes.
  • [PROMPT_INJECTION]: While the skill analyzes external code (a surface for indirect prompt injection), this is its intended primary function and no malicious patterns were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 09:20 PM
Security Audit — agent-trust-hub — vulnerability-pattern-matcher