breakdown
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs local command execution to initialize and run the rendering pipeline. It executes a setup script (
setup.sh) located in a sibling directory tree and runs a Python build script (build.py) to orchestrate the Three.js rendering process. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text labels that are interpolated into a Three.js HTML template (
template.html) usinginnerHTML. While this presents a technical surface for cross-site scripting (XSS) within the rendering environment, the risk is mitigated by the intended output format (MP4 video) and the use of an isolated headless Chrome instance for generation.
Audit Metadata