model-split-screen
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/split.pyuses thesubprocess.runfunction to executeffmpegandffprobecommands. It correctly mitigates shell injection risks by passing command arguments as a list rather than a single shell string, which is a standard security best practice for command execution. - [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it ingests untrusted external data and has the capability to write files and execute subprocesses.
- Ingestion points: The script
scripts/split.pyaccepts external input via the--left,--right,--title, and--promptarguments, which can be populated from external file contents or user-supplied strings. - Boundary markers: The instructions do not define explicit delimiters or warnings for the agent to ignore potentially malicious embedded instructions in the processed data.
- Capability inventory: The skill can execute system binaries (
ffmpeg,ffprobe) viasubprocess.runand perform file-write operations to the local filesystem. - Sanitization: Shell injection is mitigated by using argument lists in
subprocess.run. Textual input is rendered into static images using the Pillow library, preventing the content from being interpreted as executable code within the script's own logic.
Audit Metadata