Agent Router
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (LOW): The skill processes use r request s to determine routing logic withou t inpu t delimiter s. An adversaria l use r could embe d instruction s in thei r reques t to manipulate the routing decision. * Ingestion point s: Use r natura l language request s in skill trig ger s. * Boun d ary marker s: Absen t. * Capability inventor y: Interna l delegation via ACTIVATE_SKILL (low-privileg e side effect s). * Sanitization: None detected.
Audit Metadata