ad-library-teardown
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external ad libraries which could contain adversarial instructions.
- Ingestion points: External data is ingested via tools such as metaAdsAdGet, metaAdsTranscriptGet, and linkedinAdsAd referenced in SKILL.md.
- Boundary markers: The skill explicitly warns the agent in the Pitfalls section to ignore instructions inside ad copy and transcripts.
- Capability inventory: The skill is limited to data analysis and does not utilize tools for file system modification, command execution, or unauthorized network operations.
- Sanitization: Security is managed via prompt-level instructions and developer warnings provided in the skill documentation.
Audit Metadata