video-prompting-guide

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documentation includes multiple examples of shell commands for the belt CLI tool. These commands are intended to help users run video generation models with specific cinematic parameters.
  • [EXTERNAL_DOWNLOADS]: References an external installation guide hosted on the inference-sh GitHub repository. This link provides necessary setup instructions for the CLI tools described in the guide.
  • [PROMPT_INJECTION]: The skill instructions involve interpolating user-provided descriptive text into command arguments, which creates a surface for indirect prompt injection.
  • Ingestion points: User-defined video prompts in the prompt field of the input JSON structure.
  • Boundary markers: Examples utilize JSON structure to encapsulate user input, providing a basic separation between instructions and data.
  • Capability inventory: The skill uses the Bash(belt *) tool as specified in the allowed-tools metadata.
  • Sanitization: No explicit sanitization or filtering of user input is defined within the instructional text.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 02:15 AM
Security Audit — agent-trust-hub — video-prompting-guide