testing
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill defines a senior test architect persona without any instructions to override safety filters or disregard prior rules. All instructions are focused on technical quality and risk management.\n- [EXTERNAL_DOWNLOADS]: The skill recommends using reputable third-party testing tools and libraries such as k6, SonarQube, and Snyk. No suspicious remote code execution or untrusted script downloads were found.\n- [DATA_EXFILTRATION]: No sensitive file access or network exfiltration patterns were identified. Mentions of credentials and .env files are in the context of security scanning and configuration best practices.\n- [OBFUSCATION]: No encoded content, zero-width characters, or homoglyph attacks were found. The mention of zero-width characters in the references is correctly categorized as a data validation test case.\n- [INDIRECT_PROMPT_INJECTION]: The skill serves as a structural framework for testing and does not ingest or process untrusted external data that could lead to indirect injection vulnerabilities.
Audit Metadata