scroll-film
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for generating AI video prompts based on user-supplied scene descriptions and art direction, which constitutes a potential surface for indirect prompt injection.
- Ingestion points: User input for scene details and style is interpolated into templates defined in
references/chain.md. - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used in the prompt templates.
- Capability inventory: The skill uses
ffmpegvia the localscripts/verify.shscript for video analysis and directs the agent to call external AI video generation providers. - Sanitization: User input is interpolated directly into the prompt strings without explicit escaping or validation within the provided templates.
- [COMMAND_EXECUTION]: The skill includes a local bash script (
scripts/verify.sh) used for quality assurance and technical validation of video encodes and seams. - The script performs read-only operations using standard utilities like
ffprobe,ffmpeg, andawkto verify video properties and calculate SSIM scores between frames. - No arbitrary command execution or shell injection vulnerabilities were found in the script's handling of file paths.
Audit Metadata