false-positive-prompt-injection
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- SAFE (LOW): No malicious patterns detected. The presence of phrases like 'ignore previous instructions' and 'override system rules' occurs within an educational context recommending against such behaviors, rather than as an attempt to hijack the agent's control flow. The skill contains no executable code, network requests, or sensitive file access.
Audit Metadata