author-modeladvisor-checks

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a collection of documentation and templates for developing Simulink Model Advisor checks. All code patterns utilize standard MATLAB and Simulink APIs such as get_param, find_system, and ModelAdvisor.ResultDetail.
  • [PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection by processing untrusted model data (ingestion points include find_system in references/api-blocks.md and mtree in references/api-code-analysis.md) and providing modification capabilities (capability inventory includes set_param and ModelAdvisor.Action in references/api-framework.md). Boundary markers are present in the workflow requiring the agent to present a plan to the user in SKILL.md. Sanitization is implemented via mandatory static analysis using the check_matlab_code tool.
  • [COMMAND_EXECUTION]: The skill manifest references the use of platform tools check_matlab_code and evaluate_matlab_code. These are standard tools for static analysis and code evaluation within the Simulink environment and are used as intended for developer verification.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 08:23 AM
Security Audit — agent-trust-hub — author-modeladvisor-checks