video-to-superprompt
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The workflow relies on executing
ffprobeandffmpegvia the Bash tool to extract technical properties (duration, dimensions) and visual frames from video files. These operations are essential for generating the requested design analysis and are performed using standard parameters. - [EXTERNAL_DOWNLOADS]: The skill uses the
WebFetchtool to download video content from URLs provided by the user. This is a core feature enabling the analysis of remote video references as specified in the skill's goals. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted external data from videos and URLs while having filesystem and shell access.
- Ingestion points: User-supplied video files and remote URLs defined in the workflow.
- Boundary markers: The skill explicitly instructs the agent to provide its final output in a single fenced
textblock to maintain boundary integrity. - Capability inventory: Includes shell command execution via Bash and filesystem operations (
Read,Write,Edit) used for frame extraction and prompt generation. - Sanitization: While no explicit content sanitization is described, the process of extracting frames and technical metadata for analysis mitigates the risk of direct execution of malicious instructions embedded in the source media.
Audit Metadata