skills/magentosh/skills/chat-ui/Gen Agent Trust Hub

chat-ui

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads UI component definitions and registry files from the provider's official domain.
  • Evidence: npx shadcn@latest add https://ui.inference.sh/r/chat.json in SKILL.md.
  • [COMMAND_EXECUTION]: Uses the npx command to install CLI tools and related skill packages for the development environment.
  • Evidence: npx skills add belt-sh/cli and npx skills add inference-sh/skills@agent-ui in SKILL.md.
  • [PROMPT_INJECTION]: The skill creates an interface for displaying external data (chat messages), which constitutes an indirect prompt injection surface if the content is not sanitized.
  • Ingestion points: Untrusted data is ingested via the content prop of the ChatMessage component (SKILL.md).
  • Boundary markers: No delimiters or boundary markers are suggested in the provided code examples to separate instructions from data.
  • Capability inventory: The skill provides UI rendering capabilities for chat messages, input, and indicators; it does not grant system-level execution capabilities to the data being rendered.
  • Sanitization: No explicit sanitization or escaping logic is included in the usage examples, leaving implementation to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 06:04 PM
Security Audit — agent-trust-hub — chat-ui