p-video
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates the processing of untrusted user input for video generation, which constitutes a surface for indirect prompt injection.
- Ingestion points: The prompt, image, and audio fields within the --input JSON argument in SKILL.md.
- Boundary markers: Input data is encapsulated in JSON format.
- Capability inventory: The skill relies on the infsh tool to execute remote AI models on the Pruna platform.
- Sanitization: There are no explicit instructions or delimiters used to sanitize or isolate user-provided prompts from agent instructions.
- [EXTERNAL_DOWNLOADS]: Fetches installation metadata and related skill configurations from the official inference-sh GitHub repository and website.
- [COMMAND_EXECUTION]: Uses the vendor-provided infsh CLI to manage authentication and run specialized video generation applications.
Audit Metadata