marketing-xiaohongshu-specialist
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides instructions for a marketing persona. It does not include any executable code beyond a standard shell wrapper used to invoke the agent.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from social media interactions such as audience feedback, comments, and DMs. Ingestion points: SKILL.md (Phase 4, Phase 5, Learning & Memory). Boundary markers: Absent. Capability inventory: No subprocess calls, exec/eval, or network operations found in scripts. Sanitization: Absent. The risk is minimal as the skill lacks dangerous tools.
Audit Metadata