meta-ads-creative
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists solely of informational Markdown documents providing guidance on advertising strategy and copywriting. No executable code, scripts, or automated tools are included.
- [SAFE]: All external URLs point to legitimate, well-known services and platforms for advertising research, such as Meta Ad Library, TikTok Creative Center, and Google Trends. No suspicious or obfuscated links were identified.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a research phase that involves ingesting untrusted data from external sources.
- Ingestion points: Testimonials, Reddit threads, community forums, and competitor ads (documented in Phase 1 of SKILL.md).
- Boundary markers: None specified to differentiate ingested research data from system instructions.
- Capability inventory: No automated tools or code execution capabilities are defined in the skill's configuration or instructions.
- Sanitization: No sanitization or filtering methods are described for the external content retrieved during research.
Audit Metadata