skills/inference-sh-9/skills/agent-ui/Gen Agent Trust Hub

agent-ui

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches component configuration from the vendor's repository at ui.inference.sh using the npx shadcn command.
  • [EXTERNAL_DOWNLOADS]: Downloads additional agent-related UI skills from the inference-sh skills registry.
  • [PROMPT_INJECTION]: The component processes responses from an AI agent to render UI and execute tools, creating a surface for indirect prompt injection. 1. Ingestion points: The Agent component and its associated proxy route. 2. Boundary markers: No explicit markers shown in the documentation snippets. 3. Capability inventory: Includes form filling and UI scanning tools. 4. Sanitization: Implementation of human-in-the-loop approval flows serves as a control mechanism.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 05:35 PM