widgets-ui

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The documentation instructs users to install external software using npx commands, including the belt-sh/cli tool and components from ui.inference.sh.
  • [COMMAND_EXECUTION]: Provides instructions for executing shell commands that download and run code from remote registries.
  • [INDIRECT_PROMPT_INJECTION]: The skill uses agent-provided JSON to dynamically generate interactive user interfaces, creating an attack surface for UI-based deception.
  • Ingestion points: Widget and action definitions are extracted from agent responses to be processed by the React component.
  • Boundary markers: The skill lacks explicit guidance or markers to distinguish agent-generated UI from system-trusted UI within the application.
  • Capability inventory: Supports rendering interactive components such as inputs, textareas, and buttons with submit actions that can capture user data.
  • Sanitization: There is no evidence in the provided documentation of validation or sanitization of the JSON structure before it is rendered.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:23 PM
Security Audit — agent-trust-hub — widgets-ui