08-nghien-cuu-doi-thu
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional text and markdown templates for strategic planning. It contains no executable code, shell commands, or obfuscated payloads.
- [EXTERNAL_DOWNLOADS]: The instructions reference official and well-known platforms including Facebook Ads Library (mcp.facebook.com), TikTok Creative Center, and Google Trends. These are recognized services used for their intended research purposes.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from competitor websites and social media platforms. 1. Ingestion points: Competitor URLs, Facebook Ads Library output, and social media reviews. 2. Boundary markers: None explicitly defined in the provided templates. 3. Capability inventory: Generates markdown reports; no file-system write, subprocess execution, or arbitrary network capabilities identified. 4. Sanitization: None specified for the ingested text. This constitutes a standard research surface with no critical high-privilege capabilities.
Audit Metadata