media-buyer
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access techniques were detected. The skill is designed with a clear, safe workflow for building marketing campaigns.
- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it processes user-provided ad copy and image hashes. However, this is well-mitigated by the following controls:
- Ingestion points: Creative assets and copy provided via the
[IMAGE HASHES + COPY]placeholder. - Boundary markers: No explicit delimiters are used in the prompt template.
- Capability inventory: The skill utilizes a Meta Ads MCP with write access for campaign creation.
- Sanitization: No explicit sanitization of input text is performed.
- Mitigation: The skill strictly enforces a "paused-by-default" policy, requiring the user to manually review the campaign structure in a table and provide a "publish" command before any changes take effect.
Audit Metadata