mixamo-retarget
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external 3D animation files which serves as a data ingestion point.
- Ingestion points: The functions
import_fbxandimport_daeinscripts/retargeting.pyload external file data into the active Blender session. - Boundary markers: The instructions lack specific boundary markers or directives for the agent to disregard potential instructions hidden within bone names, object metadata, or file properties of the imported assets.
- Capability inventory: The skill uses the
execute_blender_codetool to interact with the Blender API and can perform file system read operations and scene modifications. - Sanitization: While
scripts/bone_matching.pyperforms regex-based normalization on bone names for algorithmic matching, it does not implement security sanitization to filter out potential prompt injection payloads.
Audit Metadata