scenario-blender-lighting-rendering
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implementation in
bx_light.pyis a legitimate utility for Blender lighting automation. Technical analysis shows that the script performs expected operations using thebpyAPI andnumpyfor numerical calculations on image data. There is no evidence of unauthorized network activity or credential harvesting. - [EXTERNAL_DOWNLOADS]:
SKILL.mdinstructs the agent to prompt the user to install sibling skills using thenpx skills add scenario-labs/skillscommand. This reference points to official resources provided by the same vendor (scenario-labs) and is a standard procedure for skill management within this ecosystem. The command is provided as a suggestion to the user and is not executed automatically by the skill. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface through the analysis of scene geometry and rendered image files. 1. Ingestion points:
BL.analyse(PNG/EXR files),BL.mask_cryptomatte(EXR manifest). 2. Boundary markers: Absent. 3. Capability inventory:bpy.ops.render.render(execution),bpy.data.images.save(file write). 4. Sanitization: Absent. However, the processing is strictly numerical (histograms, contrast, k-means clustering) and does not involve natural language parsing of the data that could trigger unintended agent behavior.
Audit Metadata