pentest-ai-llm-security
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: The skill documents common prompt injection techniques—such as instruction overrides ('Ignore previous instructions') and role-playing—specifically as test vectors for authorized security testing in SKILL.md and references/workflows.md.
- [DATA_EXFILTRATION]: The references/workflows.md file describes audit procedures for identifying sensitive information disclosure and system prompt leakage, which are standard components of the security methodology described in the skill.
Audit Metadata