short-film-editor

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Orchestrates local media processing workflows using Bash to invoke FFmpeg for video concatenation, ImageMagick for storyboard grid splitting, and Python for audio beat analysis.
  • [EXTERNAL_DOWNLOADS]: Downloads AI-generated video assets from remote URLs as part of the standard production pipeline using curl.
  • [PROMPT_INJECTION]: The skill processes user-provided narrative content to generate prompts for downstream AI video models, representing a standard surface for indirect prompt injection.
  • Ingestion points: User-supplied scripts and character attributes ingested in the SKILL.md workflow.
  • Boundary markers: Prompts are structured using headers (Style, Character, Scene), but do not employ specific adversarial instruction delimiters.
  • Capability inventory: File system management, network downloads via curl, and local command execution via Bash.
  • Sanitization: Content is interpolated into generation prompts without dedicated sanitization logic to maintain creative fidelity.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 06:03 AM
Security Audit — agent-trust-hub — short-film-editor