Skills
skills/magentosh/skills/seedance/Gen Agent Trust Hub

seedance

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted user input via the prompt parameter, creating a surface for indirect prompt injection.
  • Ingestion points: The prompt field in JSON payloads for belt app run (found in SKILL.md).
  • Boundary markers: None. Input is interpolated directly into the CLI command payload without delimiters.
  • Capability inventory: Access to Bash with the belt CLI, allowing interaction with external video generation APIs.
  • Sanitization: The skill does not implement validation or escaping for the user-provided prompt string.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute belt CLI commands. This is used to authenticate (belt login) and run video generation apps on the inference.sh platform.
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct the user to install an external dependency (belt-sh/cli) and reference installation guidelines hosted on GitHub (inference-sh/skills).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 06:04 PM
Security Audit — agent-trust-hub — seedance