blender-pipeline-automation
Warn
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [DYNAMIC_EXECUTION]: The
batch_process.pyscript contains a_modulefunction that usesimportlib.utilto dynamically load and execute Python modules from file paths provided via the--scriptcommand-line argument. This allows for the execution of arbitrary Python code within the Blender process context. - [COMMAND_EXECUTION]: The skill provides instructions and scripts for executing complex Blender CLI commands. These commands include the
-P(Python) and--python-exprflags, which enable the execution of scripts and expressions during the Blender startup process. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process a wide variety of external 3D asset files (e.g., .fbx, .glb, .blend) across entire directory trees. It lacks a mechanism to distinguish between legitimate assets and those designed to exploit the batch processing logic.
- Ingestion points: File paths and directory globs provided to the
--inargument, as well as script paths provided to the--scriptargument inbatch_process.py. - Boundary markers: None identified in the script logic or instructions to isolate processed data from the execution environment.
- Capability inventory: The skill can execute arbitrary Python scripts (
_modulefunction inbatch_process.py), perform file system operations such as writing new files (EXPORTERSmapping,bpy.ops.wm.save_as_mainfile), and utilize various Blender internal operators. - Sanitization: There is no evidence of validation or sanitization for the paths provided to the
--scriptargument or the arguments passed into the dynamically loaded scripts. - [EXTERNAL_DOWNLOADS]: The documentation in
SKILL.mdincludes instructions for downloading the Blender binary fromdownload.blender.org, which is the official distribution channel for the software.
Audit Metadata