mira-trends
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content (hashtags, sound titles, and video descriptions) from TikTok and Instagram to identify trends and generate AI video prompts. This data ingestion creates a potential attack surface for indirect prompt injection.\n
- Ingestion points: Tools
get_trendsandget_trend_videoingest metadata from external social media platforms (SKILL.md).\n - Boundary markers: The instructions lack specific delimiters or boundary markers to distinguish untrusted social media content from system instructions.\n
- Capability inventory: The skill allows for the automated creation and remaking of videos (
direct_video,generate_video) based on the fetched trending data.\n - Sanitization: While the skill provides instructions to avoid copyrighted material, it does not include technical sanitization or validation of the ingested strings to prevent them from being interpreted as commands.
Audit Metadata