saudi-arabia-grc
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements an 'Intake Gate' that requests sensitive organizational information, creating an inherent surface for indirect prompt injection where malicious data could attempt to influence advisor behavior.\n
- Ingestion points: Step 1 in
SKILL.mdprompts for detailed metadata including organization type, data classification, and cloud posture.\n - Boundary markers: The instructions lack explicit delimiters or safety instructions to isolate and ignore potentially malicious commands embedded within the user's descriptive responses.\n
- Capability inventory: The skill is strictly informational and reference-based; it does not request or utilize tools for file system modification, command execution, or network communication.\n
- Sanitization: No validation or sanitization mechanisms are defined for the content processed during the intake workflow.
Audit Metadata