security-best-practices

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill serves as a security review tool, providing guidance based on established industry standards from trusted sources such as OWASP, MDN, and official framework documentation. All external URL references target well-known, trusted domains and organizations.
  • [NO_CODE]: The skill is comprised solely of markdown files and YAML configuration. It contains no executable code (.py, .js, .sh), which significantly reduces its own attack surface.
  • [PROMPT_INJECTION]: The skill instructions define a surface for indirect prompt injection, which is necessary for its intended purpose of auditing external code.
  • Ingestion points: The agent ingests untrusted source code and repository metadata to perform security reviews as defined in the workflow (SKILL.md).
  • Boundary markers: The skill does not explicitly instruct the agent to use delimiters (like XML tags or triple backticks with specific labels) to isolate untrusted code from instructions.
  • Capability inventory: The skill has the capability to write a security report to the local file security_best_practices_report.md (SKILL.md).
  • Sanitization: There are no specific instructions regarding the sanitization or escaping of untrusted code content before it is included in the reasoning process or the final report.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 05:50 PM