chat-ui
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install UI components from a remote registry at
ui.inference.shusing theshadcnCLI tool. - [INDIRECT_PROMPT_INJECTION]: The skill defines components designed to handle and display untrusted data within a chat interface, creating a potential surface for indirect injection if inputs are not properly sanitized by the implementing application.
- Ingestion points: Untrusted data enters the context through the
ChatInputcomponent and is displayed via theChatMessagecomponent defined inSKILL.md. - Boundary markers: The provided code examples do not include explicit boundary markers or instructions to ignore embedded commands within message content.
- Capability inventory: The components are focused on UI rendering and do not include capabilities for file system access, network requests, or command execution.
- Sanitization: The documentation provides layout and styling examples but does not include internal logic for sanitizing or escaping the content processed by the chat components.
Audit Metadata