vercel-ai-sdk-expert
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill serves as a technical guide for the Vercel AI SDK, providing documentation and best practices for building AI-powered applications. It does not execute commands or access sensitive files.
- [EXTERNAL_DOWNLOADS]: The skill references standard Node.js packages from Vercel's official ecosystem, including 'ai', '@ai-sdk/react', and '@ai-sdk/openai', alongside the 'zod' validation library. These are well-known, trusted industry tools.
- [PROMPT_INJECTION]: The implementation patterns described involve processing untrusted user input, creating a surface for indirect prompt injection.
- Ingestion points: User messages are ingested via 'req.json()' in the API route example and managed through 'handleInputChange' in the React component.
- Capability inventory: The skill demonstrates tool calling functionality (e.g., 'getWeather' tool), which can be manipulated by malicious user input.
- Boundary markers: The provided code snippets do not include explicit delimiter markers or instructions to the model to ignore instructions embedded within the user data.
- Sanitization: While 'zod' is correctly used to validate structured outputs from the LLM, the examples do not show sanitization or filtering of the input prompt strings.
Audit Metadata