matlab-train-network

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a technical guide for MATLAB users to migrate from legacy neural network APIs to modern R2024a+ deep learning functions. All code examples provided use standard MATLAB syntax for training (trainnet, fitcnet), inference (minibatchpredict), and evaluation (testnet).
  • [SAFE]: No obfuscated code, hidden URLs, or malicious data exfiltration patterns were found. The skill does not perform any network operations or access sensitive system paths.
  • [SAFE]: The skill is authored by MathWorks, the vendor of MATLAB, and the instructions are consistent with official documentation. The use of MATLAB-specific tools like evaluate_matlab_code is appropriate for the stated purpose of training neural networks.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided data (e.g., tables or image datastores) for training. While this presents a theoretical attack surface for indirect prompt injection via data poisoning, it is a core functional requirement of the skill and no specific vulnerabilities were identified. Boundaries are naturally formed by the MATLAB execution environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 08:33 PM
Security Audit — agent-trust-hub — matlab-train-network