security-review

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists entirely of markdown documentation and code snippets providing security guidelines for authentication, secrets management, input validation, and cloud infrastructure. No executable scripts or tool configurations are present within the skill files.
  • [SAFE]: External references to libraries and tools (such as Zod, DOMPurify, AWS SDK, and GitHub Actions) are limited to code examples for best-practice implementations and utilize well-known, trusted industry sources.
  • [INDIRECT_PROMPT_INJECTION]: While the skill's purpose is to review user-supplied code (ingestion surface), it lacks any operational tools, network capabilities, or file-write permissions that could be leveraged by an indirect injection attack. The skill acts as a passive knowledge base for the agent.
  • [CREDENTIALS_UNSAFE]: The skill includes hardcoded secret patterns (e.g., 'sk-proj-xxxxx'); however, these are explicitly presented as 'FAIL' examples to demonstrate insecure practices and utilize generic placeholder strings that pose no risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 02:58 PM