web-error-handling-error-boundaries
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides implementation patterns that render
error.messagedirectly into the UI (e.g., inErrorFallbackandRetryLimitedBoundarycomponents). If an error message originates from an untrusted source, such as a malicious API response or manipulated data, it could contain instructions intended to influence an agent that reads the UI content. - Ingestion points: The
error.messageproperty is interpolated into the DOM in multiple components acrossexamples/core.mdandexamples/recovery.md. - Boundary markers: No specific delimiters or instructions to ignore embedded content are included in the provided UI templates.
- Capability inventory: The skill focus is UI components, but an agent implementing these patterns may have access to tools for file operations or network requests.
- Sanitization: While React's default rendering provides protection against XSS, it does not sanitize the semantic content of the strings against indirect prompt injection.
Audit Metadata