wstg-security-testing

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a legitimate security auditing framework using the established OWASP WSTG methodology for educational and professional use.
  • [DATA_EXFILTRATION]: Analysis of reference/PAYLOADS.md confirms that sensitive file paths (e.g., /etc/passwd) and metadata service URLs (e.g., 169.254.169.254) are documented as instructional payloads for testing external targets. No evidence was found of these being used to exfiltrate data from the local agent environment.
  • [COMMAND_EXECUTION]: The Python scripts scripts/wstg_checklist.py and scripts/wstg_lookup.py use standard libraries for processing local JSON data and generating markdown reports. They do not utilize unsafe subprocess calls or dynamic code execution on user-provided input.
  • [PROMPT_INJECTION]: The skill analyzes untrusted data such as codebase diffs and PR descriptions as part of its auditing workflow.
  • Ingestion points: Source code and PR diff review modes defined in SKILL.md.
  • Boundary markers: No explicit delimiters are used in the scripts to isolate untrusted data.
  • Capability inventory: Execution of local utility scripts for checklist scoring and ID lookup.
  • Sanitization: Relies on the host model's safety constraints as no explicit sanitization of reviewed code is performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 02:42 PM
Security Audit — agent-trust-hub — wstg-security-testing