The Agent Skills Directory

[SAFE]: The skill instructions exclusively use the official Blender Python API (bpy) to manage compositing nodes, render settings, and file outputs. No suspicious or hidden logic was detected.
[DATA_EXPOSURE]: File write operations are restricted to standard temporary directories (/tmp/comp_output/) for rendering purposes, and file reads are limited to project-relative paths (e.g., //plates/rooftop_plate.jpg). No access to sensitive system files or credentials was found.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns or downloads from untrusted sources are present. The repository reference in the metadata points to the official Blender project.
[OBFUSCATION]: The content is presented in clear, readable Python and Markdown. There is no evidence of Base64 encoding, zero-width characters, homoglyphs, or other obfuscation techniques.
[COMMAND_EXECUTION]: Shell command examples are limited to standard Blender CLI usage (blender --background --python) for executing the provided scripts in a headless environment.
[INDIRECT_PROMPT_INJECTION]: The skill does not ingest untrusted natural language data that could be used for prompt injection attacks. It primarily processes image data and node configurations within the Blender environment.

blender-compositing