coreml
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides legitimate technical documentation and implementation rules for Apple's Core ML framework, including Vision and Natural Language integration.
- [SAFE]: Best practices for security and performance are correctly identified, such as performing expensive model operations off the main thread and validating input feature names and shapes to prevent runtime crashes.
- [SAFE]: The mention of runtime model downloading and compilation is framed as a standard solution for handling large or dynamic models using official Apple APIs (URLSession, MLModel.compileModel).
- [SAFE]: No patterns of prompt injection, data exfiltration, obfuscation, or malicious code execution were detected across the provided files.
Audit Metadata