long-to-shorts
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted transcript data and visual metadata, creating an attack surface where instructions embedded in the content could attempt to influence the agent's behavior.
- Ingestion points: Transcript content is read for segment identification in SKILL.md; visual context metadata (semanticProfile.visualContext) is inspected for reframing logic.
- Boundary markers: Absent. The instructions do not define specific delimiters or directives to ignore instructions found within the processed content.
- Capability inventory: Capabilities are limited to media processing commands including cutRanges, trimClip, reframeAuto, addTitle, setAudio, syncAudioToVideo, createMulticamGroup, and switchAngle.
- Sanitization: Absent. The skill does not specify validation or sanitization of transcript text prior to processing.
Audit Metadata