blender-retopology

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is a specialized technical resource for Blender retopology. It implements advanced geometric algorithms (geodesic fields, radial density warps, 3-to-1 quad reductions) that align with the provided expert references from the Blender Studio and other industry professionals.
  • [DATA_EXPOSURE]: The script bx_retopo.py includes functions for rendering diagnostic images (face_set_review) and exporting finished assets (handoff). These operations are standard utilities for 3D workflows and are restricted to local file system paths specified by the user or common temporary directories, posing no risk of unauthorized data exfiltration.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external mesh data (sculpts, scans, or AI-generated geometry) to generate new topology. While the skill ingests untrusted geometric data, the operations are confined to geometric transformations and standard Blender APIs. There is no evidence of instructions within the data being interpreted as agent commands.
  • [METADATA_POISONING]: The skill contains local development paths and specific author metadata (e.g., paths under /Users/emmanuel/). These are benign artifacts of the skill's development and testing environment and do not impact the security or integrity of the execution.
  • [UNVERIFIABLE_DEPENDENCIES]: The script imports bx_audit and bx_review, which appear to be part of the larger 'Blender Expert Skills' suite mentioned in the documentation. While these files are not provided in this specific package, their usage in the code is consistent with the described utility purposes (topology auditing and review renders).
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 05:46 AM