solidcraft-modeling

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a standard set of instructions for using a specialized CAD modeling tool via an MCP server. The operations described (modeling, measuring, screenshotting, and exporting) are consistent with the skill's stated purpose of 3D solid modeling.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of user-defined parameters and mathematical expressions for CAD operations.
  • Ingestion points: Untrusted data enters the context through user-provided parameter values and mathematical expressions (e.g., in parameters.add and sketch.dimension calls).
  • Boundary markers: None explicitly defined in the instructions for separating user data from system commands.
  • Capability inventory: The skill utilizes tools for design modification (batch, execute), design state inspection (measure, body_topology, inspect_design), visual verification (screenshot), and file persistence (export, save).
  • Sanitization: The skill relies on the SolidCraft backend to parse and validate mathematical expressions, which is a standard pattern for CAD tools.
  • [DATA_EXFILTRATION]: The skill includes functionality to export design files (e.g., .step, .stl) and take screenshots for verification. These are legitimate features of a modeling environment and are not used here for unauthorized data transmission.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 05:21 PM