foundation-models
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill includes detailed instructions on how to prevent prompt injection by distinguishing between developer-defined instructions and user-controlled prompts. It also provides guidance on handling platform-level guardrail violations.
- [CREDENTIALS_UNSAFE]: The documentation incorporates proactive safety advice, explicitly warning against hardcoding API keys and recommending the use of the Apple Keychain for managing authentication tokens for third-party models.
- [EXTERNAL_DOWNLOADS]: The skill references official resources from trusted organizations, including the
apple/foundation-models-utilitiesrepository on GitHub and Swift packages from Google and Anthropic for model integration.
Audit Metadata