components
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill implements a system where an AI agent can dynamically generate and update UI component properties (generative and interactable components). This design creates an indirect prompt injection surface where the AI's output, used to populate component props, could be influenced by untrusted input.\n
- Ingestion points: SKILL.md (via AI processing of user messages or external data to satisfy
propsSchema).\n - Boundary markers: None explicitly mentioned in the code examples.\n
- Capability inventory: AI can set props for React components which are subsequently rendered in the user interface.\n
- Sanitization: Relies on React's default escaping for string children and developer-defined Zod schemas.\n- [EXTERNAL_DOWNLOADS]: The skill references '@tambo-ai/react' and 'zod'. As '@tambo-ai/react' is a resource from the known vendor 'tambo-ai' and 'zod' is a standard library, these are documented as safe dependencies.
Audit Metadata