video-design

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill provides markdown-based guidelines and templates for video design workflows. It does not include any executable scripts, shell commands, or network operations.
  • [EXTERNAL_DOWNLOADS]: The skill references external design resources and repositories (e.g., Leonxlnx/taste-skill, anthropics/skills). These are documented as design sources for the user to consider installing manually via plugin commands; the skill itself does not perform any automatic downloads or execution of remote code. References to these well-known repositories are informational.
  • [COMMAND_EXECUTION]: The documentation includes example shell commands (e.g., uv run tools/qc.py) that the user is expected to run locally to validate their video projects. These commands interact with standard project tools and do not represent a security risk.
  • [INDIRECT_PROMPT_INJECTION]: While the skill processes project metadata like project.json and design.md to establish design context, it does not ingest untrusted third-party data that could lead to prompt injection. The workflow is entirely user-driven and project-local.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 11:37 PM