skills/yodaos-project/aiui/aiui-dev/Gen Agent Trust Hub

aiui-dev

Pass

Audited by Gen Agent Trust Hub on Nov 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill consists primarily of Markdown documentation and boilerplate assets designed to guide an AI agent in developing and debugging AIUI applications. No malicious patterns, such as prompt injection, credential theft, or unauthorized remote code execution, were detected.
  • [INDIRECT_PROMPT_INJECTION]: The documentation for the a2ui component (found in references/components/a2ui.md) describes a system where an AI agent generates UI structures at runtime via JSON commands. This represents a vulnerability surface for indirect prompt injection if the agent incorporates untrusted external data into the generated UI instructions.
  • Ingestion points: The a2ui component accepts JSON instructions through the commands property and the context.write() JavaScript method.
  • Boundary markers: The provided documentation does not define specific delimiters or instructions to prevent the agent from obeying commands embedded within user-supplied data during UI generation.
  • Capability inventory: Applications built with this skill have access to network requests (fetch, WebSocket), persistent storage, and hardware sensors (GPS, microphone, Bluetooth) as detailed in the references/apis/ folder.
  • Sanitization: The reference materials do not describe explicit sanitization or validation protocols for the JSON command streams used to generate the user interface.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 20, 2026, 05:41 AM