The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides patterns for opening and batch processing external .blend files, creating a surface where untrusted data could potentially influence agent behavior through malicious metadata or object names.
Ingestion points: File reading occurs via bpy.ops.wm.open_mainfile in Section 6 and Example 2.
Boundary markers: Absent; there are no instructions provided to validate file integrity or ignore embedded data instructions.
Capability inventory: The skill provides instructions for file system writes (bpy.ops.wm.save_as_mainfile, bpy.ops.wm.obj_export, os.makedirs), object manipulation, and scene cleanup.
Sanitization: Absent; the example scripts do not implement sanitization for file paths or object properties before processing.
[COMMAND_EXECUTION]: The skill instructs the agent on how to execute the Blender binary from the command line using the --python and --background flags. This is the core functionality of the skill and uses standard application parameters.

blender-scripting