video-ad-analysis
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No instructions to bypass safety filters or override system prompts were found.- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or access to sensitive local files (e.g., SSH keys, AWS configs) were detected.- [Remote Code Execution] (SAFE): No commands to download and execute remote scripts or install unauthorized packages were found.- [Indirect Prompt Injection] (SAFE): The skill ingests untrusted data from video URLs or descriptions (Ingestion Point). It lacks explicit boundary markers (Boundary Markers). However, the skill possesses no high-risk capabilities like subprocess calls, file writing, or network operations (Capability Inventory), and there is no evidence of sanitization (Sanitization). Due to the lack of exploitable capabilities, this is considered safe.
Audit Metadata