analyze-footage
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill interacts with TwelveLabs, an established video AI platform, to upload and analyze video assets as part of its core video-understanding workflow.
- [INDIRECT_PROMPT_INJECTION]: The skill processes descriptive data derived from external video files, creating a potential surface for indirect prompt injection if the video content contains adversarial instructions.
- Ingestion points: Video analysis results and scene descriptions retrieved from the
async-analyse-videoandsync-analyse-videotools inSKILL.md. - Boundary markers: The skill does not employ explicit delimiters or system instructions to distinguish analyzed metadata from authoritative agent directives.
- Capability inventory: File system inspection via
ls, asset creation (create-asset), asset deletion (delete-asset), and search indexing capabilities. - Sanitization: Metadata and descriptions returned from the TwelveLabs API are processed and presented without intervening sanitization or validation steps.
Audit Metadata