ads-audit
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data in the form of account exports, reports, and screenshots from multiple advertising platforms, creating a potential surface for indirect prompt injection.
- Ingestion points: User-provided data including Google Ads exports/reports, Meta Ads Manager exports/screenshots, LinkedIn Campaign Manager exports, TikTok Ads Manager exports, and Microsoft Ads account exports (SKILL.md).
- Boundary markers: Not present. There are no instructions to the agent to ignore or delimit embedded instructions within the processed files or images.
- Capability inventory: The skill generates multiple output files (ADS-AUDIT-REPORT.md, ADS-ACTION-PLAN.md, ADS-QUICK-WINS.md) and delegates tasks to subagents (audit-google, audit-meta, etc.), which could be influenced by malicious content in the input data.
- Sanitization: Not present. No explicit sanitization or validation of the input data structure or content is mentioned.
Audit Metadata