video-to-superprompt
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The main instruction file provides shell commands for
ffmpegandffprobeto perform technical video analysis. These commands utilize a$VIDEOvariable for file paths; while standard for the task, this pattern relies on the execution environment to sanitize filenames to prevent command injection. - [INDIRECT_PROMPT_INJECTION]: The skill is built to process and analyze external, untrusted content such as video files, URLs, and web source code, which serves as a surface for indirect prompt injection attacks.
- Ingestion points: External video files (local or remote paths), URLs, and associated HTML/CSS/JS source code as defined in the workflow.
- Boundary markers: The instructions lack specific boundary markers or warnings to the agent to disregard instructions that might be embedded within the analyzed content.
- Capability inventory: The skill uses shell commands (
ffmpeg,ffprobe,mkdir) to interact with the filesystem and can read files within the local repository or workspace. - Sanitization: There are no procedures defined to sanitize or validate the content extracted from the videos or source code before the agent processes it.
Audit Metadata