handraw-video-producer
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied assets, project data, and configurations to drive automated rendering engines and UI automation. To mitigate risks from untrusted content, the skill implements a strict production-stage architecture (verified in
production_stage.py) that demands human verification of the project plan and verification of individual assets before synthesis is permitted. - [COMMAND_EXECUTION]: Automated execution of CLI tools (FFmpeg, Node.js) and package managers (npm, pip) is utilized for media encoding and runtime dependency management. The skill manages these through
runtime_tools.py, which provides controlled resolution of system binaries and allows for isolated execution of production tasks. - [DYNAMIC_EXECUTION]: Runtime code assembly is used for animation rendering and vendor module management, such as the Remotion bundling process in
lib/shot_engines.pyand dynamic loading inlib/jianying_backend.py. This functionality is essential for the skill's primary purpose and is protected by structural design and manual oversight checkpoints.
Audit Metadata