ads-monitor
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (normalized snapshots), creating a potential surface for indirect prompt injection. Ingestion Point: Step 1 (loading snapshots). Boundary Markers: None. Capability Inventory: No executable code or tools defined. Sanitization: None.
- [NO_CODE]: The skill contains only markdown instructions and no executable scripts, which limits the potential for direct malicious behavior.
Audit Metadata