ai-product-development
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive guidelines for secure AI product development. It explicitly instructs developers to treat model-generated tool arguments as untrusted and mandates server-side re-authorization and validation at the time of execution.
- [SAFE]: The implementation examples in
references/streaming.mddemonstrate secure coding practices, including schema-based request validation using Zod and the proper handling of environment variables for sensitive provider configuration. - [SAFE]: All identified Node.js dependencies are established, well-known libraries from trusted organizations (Vercel and Anthropic), and are used appropriately within the provided code context.
Audit Metadata