vfx-review
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests Studio screenshots and reference images to be analyzed by a sub-agent. Ingestion points include screenshots collected via script and goal board images. Boundary markers are present in the form of a structured prompt for the sub-agent. Capabilities include file read/write operations and subprocess execution for media tools like ffmpeg. Sanitization is inherent in the conversion of data into visual strips and grids for side-by-side comparison.
- [DYNAMIC_EXECUTION]: The skill requires the agent to generate local Python scripts for file management and image grid creation, such as collect_captures.py and board.py. This is a routine activity for the skill's primary purpose of automating the VFX review process and does not involve executing untrusted remote code.
Audit Metadata