The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run and subprocess.Popen to execute external binaries including Blender, FFmpeg, and Real-ESRGAN across multiple scripts (scripts/finish.py, scripts/prep_product.py, scripts/render_shots.py). These calls rely on environment variables and the system PATH to locate executables, which is a necessary part of the skill's media processing pipeline.
[DYNAMIC_EXECUTION]: The script scripts/pfilm.py uses the __import__ function to dynamically load the mathutils module for matrix operations within the Blender environment. This is a standard programmatic pattern for Blender-specific scripts but represents a dynamic module loading capability.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including user-provided product photos and JSON shot specifications, which influence the rendering and editing logic.
Ingestion points: scripts/prep_product.py accepts a photo file path as a command-line argument; scripts/pfilm.py, scripts/render_shots.py, and scripts/finish.py ingest JSON configuration files for shot and edit sequences.
Boundary markers: No specific boundary markers or explicit sanitization warnings for embedded instructions are present in the skill's documentation.
Capability inventory: The skill possesses the capability to execute shell commands via subprocess, perform file system writes for rendering frames, and load Python modules dynamically.
Sanitization: CLI arguments are handled via argparse and data structures are parsed via json.load, but the skill does not implement specific sanitization of image metadata or content validation for the JSON fields used to drive the rendering engine.

blender-product-film