linkedin-media-prep
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on several shell-based tools for media processing to fulfill its primary purpose.
- Instructions include usage of
magick(ImageMagick),ffmpeg,ffprobe, andsips(macOS native tool). - These tools are used for operations such as cropping, resizing, padding, and transcoding media files locally.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted media files provided by the user, which represents a potential indirect injection attack surface.
- Ingestion points: Media files provided as
<input>(referenced throughoutSKILL.md). - Boundary markers: The instructions consistently use double quotes (e.g.,
"<input>") when templating shell commands, providing basic mitigation against simple filename-based command injection. - Capability inventory: The skill utilizes significant file system interaction and media processing capabilities via shell command execution as documented in
SKILL.md. - Sanitization: The skill assumes the execution environment correctly handles and escapes user-supplied filenames when populating the provided command templates.
Audit Metadata