The Agent Skills Directory

[SAFE]: The skill utilizes a dedicated Python module (scripts/bx_retopo.py) to perform geometric operations using Blender's bpy and bmesh APIs. All operations are local to the Blender session and no unauthorized network or system access was identified.
[SAFE]: The skill references other skills from the same author (scenario-labs) and points to well-known educational platforms like YouTube for reference material, which are trusted sources in this context.
[INDIRECT_PROMPT_INJECTION]: The skill processes external mesh data (sculpts) which constitutes an ingestion point for potentially malicious instructions hidden in geometric data or metadata. Ingestion points: Untrusted mesh objects (sculpt) are passed to various analysis and processing functions in scripts/bx_retopo.py. Boundary markers: The workflow in SKILL.md defines several 'GATE' points that require the agent to inspect and map the data before proceeding with geometry generation. Capability inventory: The toolkit includes extensive capabilities for modifying geometry, creating objects, and generating render sheets for review. Sanitization: The skill implements automated 'Quality gates' and 'topology_report' functions that enforce technical constraints (e.g., manifoldness, vertex valence, quads percentage), which effectively sanitize the input by ensuring it adheres to strict geometric standards.

scenario-blender-retopology