skills/qu-skills/skills/seedance/Gen Agent Trust Hub

seedance

Pass

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the belt CLI tool to execute video generation commands through the authorized Bash tool. This functionality is consistent with the skill's stated purpose of interfacing with the Seedance API.\n- [EXTERNAL_DOWNLOADS]: The skill provides links to official installation guides and documentation hosted on inference.sh and its GitHub repository. These resources are part of the legitimate service infrastructure and do not pose a security risk.\n- [PROMPT_INJECTION]: The skill defines a surface where user inputs like prompts and media URLs are interpolated into shell commands. While this creates a potential surface for indirect injection, the examples use structured JSON arguments to mitigate typical command injection risks. Ingestion points include prompt, image, and reference_images parameters in SKILL.md. Boundary markers are present in the form of JSON object encapsulation within CLI arguments. Capability inventory includes the Bash tool. Sanitization is not explicitly handled in the markdown text; the executing agent is responsible for validating inputs before tool execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 12, 2026, 07:37 PM
Security Audit — agent-trust-hub — seedance