music-montage
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external media files and audio content to extract metadata such as beats and lyrics for video editing purposes.
- Ingestion points: Media content retrieved via
fetchStockMedia,generateMedia, and user-supplied files processed byaudio_inspect. - Capability inventory: File manipulation and media editing tools including
sliceClip,addMarkers,setKeyframeTrack, andaddTitle. - Boundary markers: The instructions do not define specific delimiters to separate untrusted media metadata (like lyrics) from the agent's instructional context.
- Sanitization: There is no mention of filtering or sanitizing text extracted from audio (lyrics) before it is used in tools like
addTitleoraddCaptions.
Audit Metadata