skills/inference-sh/skills/p-video/Gen Agent Trust Hub

p-video

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the 'belt' CLI tool via Bash to interact with remote inference services. Execution is limited by the allowed-tools configuration to the 'belt' command space.
  • [EXTERNAL_DOWNLOADS]: The skill references installation scripts and additional skill components hosted on the official GitHub repository for inference.sh.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided prompts and media URLs which are interpolated into command arguments for video generation.
  • Ingestion points: User-defined text and URLs provided as JSON inputs to the belt app run command in SKILL.md.
  • Boundary markers: Input parameters are structured within JSON objects.
  • Capability inventory: The skill can execute the belt CLI tool to initiate remote video generation tasks.
  • Sanitization: No specific sanitization logic is provided in the instructions; input safety relies on the downstream inference platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 10:35 AM
Security Audit — agent-trust-hub — p-video