The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted external data in the form of reference images and texture atlases provided by the user.
Ingestion points: Image files are read by scripts/template_analyzer.py and scripts/silhouette_validator.py using the OpenCV library.
Boundary markers: No explicit boundary markers or instructions are provided to the agent to ignore potentially malicious metadata or steganographically embedded instructions within the source images.
Capability inventory: The skill utilizes Bash to run local Python analysis scripts and the mcp__blender__execute_blender_code tool to manipulate the 3D environment.
Sanitization: No sanitization of image metadata is performed; the skill relies on standard image decoding via cv2.imread.
[COMMAND_EXECUTION]: The instructions direct the agent to execute local Python scripts (template_analyzer.py and silhouette_validator.py) using the Bash tool to perform component extraction and mask validation.
[DYNAMIC_EXECUTION]: The skill generates and executes Blender Python code at runtime to build 3D geometry from extracted image contours, which is the primary intended function of the skill.

reference-to-3d