ffmpeg-core

Pass

Audited by Gen Agent Trust Hub on Oct 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation binaries and official documentation from ffmpeg.org, which is the official well-known domain for the project.
  • [COMMAND_EXECUTION]: The skill provides numerous examples of shell commands involving ffmpeg and ffprobe for media processing, such as format conversion, scaling, and metadata extraction.
  • [PRIVILEGE_ESCALATION]: Documentation for installing the tool on Linux-based systems includes the use of sudo for package management (e.g., sudo apt update && sudo apt install ffmpeg).
  • [INDIRECT_PROMPT_INJECTION]: The skill describes an attack surface where filenames or list files (like filelist.txt) are used as arguments in shell commands. If an agent constructs these commands using unsanitized external input, it could lead to command injection vulnerabilities.
  • Ingestion points: File paths and input filenames used in ffmpeg and ffprobe templates within SKILL.md.
  • Boundary markers: Absent; command templates do not include delimiters or warnings to ignore embedded instructions in input data.
  • Capability inventory: Shell command execution via ffmpeg, ffprobe, brew, apt, dnf, and choco.
  • Sanitization: Absent; the skill does not provide instructions for escaping or validating external file paths or contents.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 3, 2026, 08:10 PM