subtitle-burner

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/burn.py invokes ffmpeg and ffprobe via subprocess.run. It utilizes list-based arguments, which is a secure practice to prevent shell injection vulnerabilities.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect injection via its input parameters. Specifically, the --font, --primary-colour, and --outline-colour arguments are interpolated into the FFmpeg subtitles filter string without escaping.
  • Ingestion points: User-provided inputs for font styles and colors in scripts/burn.py (referenced in SKILL.md).
  • Boundary markers: The script provides specialized escaping for the subtitle file path, but lacks delimiters or escaping for the force_style component values.
  • Capability inventory: File system writes (output MP4) and process control (executing FFmpeg) as defined in SKILL.md metadata.
  • Sanitization: Missing validation for string-based style arguments, allowing potential manipulation of the FFmpeg filter graph syntax.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 08:36 PM
Security Audit — agent-trust-hub — subtitle-burner