ugc-ad-scripts
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, such as web-scraped customer reviews and support ticket exports, which could contain malicious instructions aimed at overriding agent behavior.
- Ingestion points: Customer reviews pulled from the web, support ticket data, and sales-call notes (SKILL.md, Workflow Step 2).
- Boundary markers: Absent. The skill does not provide instructions to use delimiters or protective wrappers when processing external content.
- Capability inventory: None. The skill consists exclusively of markdown instructions and references; it does not include any scripts, executables, or access to high-privilege system tools.
- Sanitization: Absent. The workflow explicitly instructs the agent to mirror customer language back into the output without a filtering or sanitization phase.
- [NO_CODE]: The skill does not contain any executable scripts, binary files, or automated installation procedures, limiting its execution environment to text generation within the agent's standard interface.
Audit Metadata