matlab-integrate-pytorch-vision
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill is designed to fetch Python packages from PyPI and clone repositories from GitHub using the
MPyReqAPI. Examples provided in the reference files target well-known research repositories and official organizations including Apple and Meta. - [COMMAND_EXECUTION]: Generated MATLAB code performs local file system operations using
rmdiranddeleteto manage the installation environment and resolve dependency conflicts (e.g., removing incompatiblenumpyversions). Additionally, the skill utilizessavepathto persistently update the MATLAB path during initial setup of the prerequisiteMPyReqlibrary. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it analyzes user-provided GitHub repositories to generate MATLAB code.
- Ingestion points: Analyzes external files such as
README.md,setup.py, andrequirements.txtfrom remote repositories to determine installation and inference logic. - Boundary markers: Absent; there are no specified delimiters or warnings to ignore embedded instructions within the analyzed repository data.
- Capability inventory: The skill uses MATLAB MCP tools to execute generated code for verification and produces scripts capable of network access and file system modification.
- Sanitization: Absent; the agent is directed to translate logic directly from the repository content into executable MATLAB code.
Audit Metadata