meta-ads-creative

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill files were audited for malicious patterns, including prompt injection, data exfiltration, and persistence mechanisms. No such patterns were found. The skill is purely educational and instructional.
  • [NO_CODE]: No programming code (Python, JavaScript, Shell) is included in the skill package. All files are Markdown-based documentation and ad-copy frameworks.
  • [SAFE]: The research phase describes the collection of data from external sources such as customer feedback and social media (Reddit, Meta Ad Library). This is documented as a manual research process with no automated ingestion or processing scripts that could be exploited.
  • Ingestion points: Testimonials and external research platforms (SKILL.md, creative-research-methods.md).
  • Boundary markers: Not applicable as the process is described for manual human or agent execution.
  • Capability inventory: None. The skill possesses no capabilities for subprocess execution, network communication, or filesystem modification.
  • Sanitization: Not applicable for this non-executable skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 12:38 AM