trend-scan
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from social media platforms (TikTok, YouTube, Reddit, etc.) that could contain adversarial instructions.
- Ingestion points: Social media content retrieved through platform search and hashtag tools (SKILL.md).
- Boundary markers: The instructions lack explicit markers or warnings to isolate retrieved content from the agent's instructional context.
- Capability inventory: The agent uses specialized ScrapeAtlas CLI tools for social media data collection.
- Sanitization: No sanitization or filtering is required for the external content before processing.
Audit Metadata