foundation-models-on-device
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as technical documentation for developers implementing on-device AI features. It covers text generation, structured output, and tool calling using futuristic placeholder APIs (iOS 26).
- [SAFE]: The content emphasizes user privacy and offline support, ensuring that AI operations are performed locally on the user's device rather than in the cloud.
- [SAFE]: Code snippets provided in Swift demonstrate standard programming practices for interacting with on-device models and do not contain any harmful commands, remote execution, or data exfiltration logic.
- [SAFE]: The skill describes an architecture for processing user input and invoking local tools (e.g., RecipeSearchTool). While this represents a standard data ingestion surface, the documentation includes best practices for using system instructions to enforce safety constraints.
Audit Metadata