video-preview
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends the use of
uvx videoclaw, which results in the download and execution of the 'videoclaw' package from the PyPI registry. - [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the 'videoclaw' utility to perform previews of images and videos.
- [INDIRECT_PROMPT_INJECTION]: The skill accepts a
<file_path>argument that is interpolated into a shell command, representing a surface for processing potentially untrusted file data. -
- Ingestion points: The
<file_path>parameter in the preview command within SKILL.md.
- Ingestion points: The
-
- Boundary markers: None defined to isolate the user input.
-
- Capability inventory: Execution of shell commands via the
videoclawtool in SKILL.md.
- Capability inventory: Execution of shell commands via the
-
- Sanitization: No explicit validation or sanitization of the input path is specified.
Audit Metadata