tanstack-ai-core-chat-experience
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill documents the construction of a chat interface that processes user-controlled message arrays through an LLM, representing a standard surface for indirect prompt injection.
- Ingestion points: User input is accepted via the
messagesarray in the request body of the/api/chatroute. - Boundary markers: The setup utilizes standard
systemPromptsfor defining agent behavior but does not implement specific delimiters for user-provided data within the message parts. - Capability inventory: The skill facilitates tool execution through Model Context Protocol (MCP) clients and browser-based audio/image handling.
- Sanitization: The demonstrated code patterns rely on the underlying AI provider's safety filters rather than manual input sanitization.
Audit Metadata