polyhaven-studio-setup

Warn

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The script constructs and sends Python code to a local Blender MCP server by interpolating command-line arguments directly into strings.
  • File: scripts/polyhaven-studio-setup.js
  • Vulnerability: The variable matName, which can be assigned the raw string value from the --ground argument, is interpolated into a Python script passed to Blender's execution engine: bpy.data.materials.get("${matName}").
  • Evidence: res = await send('execute_code', { code: ... bpy.data.materials.get("${matName}") ... } on lines 105 and 112.
  • Risk: If an attacker can influence the arguments passed to this script (e.g., through a manipulated user request to the AI agent), they can execute arbitrary Python commands in the Blender process by providing a payload such as "); import os; os.system('...'); #.
  • Mitigation: Sanitize all input arguments before interpolation. Use a whitelist for allowed texture IDs or escape double quotes in the input string.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 15, 2026, 11:50 PM
Security Audit — agent-trust-hub — polyhaven-studio-setup