The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/mx_light.py utilizes subprocess.run to invoke the Arnold kick utility and the Maya Render command. These operations are essential for its functionality as a rendering automation tool. The command arguments are constructed using validated internal functions and target local file paths.
[DYNAMIC_EXECUTION]: The skill uses __import__ in its probe() function to verify the presence of environmental dependencies such as numpy, PIL, and OpenImageIO. Additionally, it invokes maya.mel.eval to execute the arnoldRender command; this is the standard method for triggering Maya's internal MEL-based processes from a Python context.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface via Maya scene files, EXR/PNG image metadata, and JSON render briefs. This data is used exclusively for mathematical calculations (e.g., luminance, saliency) and render configuration. There are no mechanisms for external data to be interpreted as agent-level instructions.
[SAFE]: All external references, including sibling skills and documentation URLs, originate from trusted or vendor-aligned sources. No evidence of data exfiltration, credential harvesting, or malicious obfuscation was found.

scenario-maya-lighting-rendering