ai-video-generation
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and requires external resources for installation and extended functionality.
- References installation guidelines hosted on a public GitHub repository ('inference-sh/skills').
- Recommends the installation of multiple external skill packages via the NPM registry under the 'belt-sh' and 'inference-sh' organizations.
- [COMMAND_EXECUTION]: The skill is centered around executing shell commands via the 'belt' CLI tool.
- The instructions provide patterns for authentication ('belt login') and application execution ('belt app run') to interact with remote AI models.
- [INDIRECT_PROMPT_INJECTION]: The skill processes remote data which could potentially contain malicious instructions intended to influence the agent's behavior.
- Ingestion points: The skill provides numerous examples in 'SKILL.md' that accept external URLs for images, audio, and video files (e.g., 'image_url', 'audio_url', 'video_url').
- Boundary markers: There are no explicit boundary markers or instructions telling the agent to ignore potentially embedded data or metadata in the processed files.
- Capability inventory: The skill has the capability to execute shell commands ('belt') which could be leveraged if the agent is misled by data content.
- Sanitization: No sanitization or integrity checks for the remote URLs or the media content are defined within the skill instructions.
Audit Metadata