ffmpeg
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions provide templates for shell commands that are intended to be populated with user-supplied data (such as filenames and encoding parameters).
- Ingestion points: User-provided filenames and parameters are interpolated into
ffmpegandffprobecommand strings (SKILL.md). - Boundary markers: None identified; there are no explicit instructions to the agent to treat user-provided strings as untrusted or to use delimiters to prevent command injection.
- Capability inventory: The skill uses shell command execution to perform media conversion and inspection (SKILL.md).
- Sanitization: No sanitization, escaping, or validation logic is provided in the command templates to handle potentially malicious input strings.
Audit Metadata