forgent3d-cad

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The file bin/cad-build is a launcher script that uses node:child_process.spawn to execute the @forgent3d/cad-build CLI tool or a local binary specified by an environment variable.
  • [EXTERNAL_DOWNLOADS]: The launcher script is configured to download and execute the latest version of the @forgent3d/cad-build package from the npm registry using npx. This package is associated with the skill author's organization.
  • [REMOTE_CODE_EXECUTION]: The skill instructions in SKILL.md direct the agent to use cad-build run-python --code '...' to execute dynamically generated Python snippets within the model's namespace. This is used to perform measurements that are not included in the standard build report.
  • [DATA_EXFILTRATION]: The cad-build preview command generates a hosted 3D preview link. The instructions state that this process involves model data being accessible via a Forgent3D account login.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it requires the agent to interpret data from model.py and the diagnostic output of the cad-build tool.
  • Ingestion points: The agent reads the contents of model.py and the dimension/diagnostic reports returned by the cad-build rebuild command.
  • Boundary markers: No specific delimiters or warnings to ignore instructions within the processed data are provided.
  • Capability inventory: The agent can execute shell commands via the bin/cad-build wrapper and run arbitrary Python code via the run-python command.
  • Sanitization: There is no evidence of sanitization or filtering applied to the tool outputs or the model code before the agent processes them.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 06:47 AM