FFmpeg Video Editor

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to generate functional bash commands for video processing. While the agent does not execute them directly, the output is specifically formatted for shell execution.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) through user-supplied filenames and parameters.
  • Ingestion points: User requests containing filenames, timestamps, and processing options (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions to ignore instructions embedded in filenames are provided.
  • Capability inventory: Generation of FFmpeg commands with potential read/write access to the local file system.
  • Sanitization: The skill does not define any sanitization or escaping rules for the interpolated user input, relying only on double quotes in examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 05:13 AM