Skills
▥
skills/jasonjgardner/blockbench-mcp-project/blockbench-mcp-overview/Gen Agent Trust Hub

blockbench-mcp-overview

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (MEDIUM): The documentation defines a risky_eval tool specifically for executing code directly within the Blockbench application (e.g., risky_eval: code="Cube.all.length"). This capability allows for dynamic execution of arbitrary logic.
  • [PROMPT_INJECTION] (LOW): The skill functions as an interface for processing external 3D model data, textures, and Hytale assets. This creates an indirect prompt injection surface where malicious instructions embedded in model metadata could potentially influence the code passed to the risky_eval tool.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:44 PM