NYC

miniprogram-development

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Command Execution] (LOW): The skill provides instructions for the agent to execute CLI commands to open the WeChat Developer Tools. The paths provided (C:\Program Files (x86)\Tencent\微信web开发者工具\cli.bat and /Applications/wechatwebdevtools.app/Contents/MacOS/cli) are standard installation directories for the official Tencent software.
  • [External Downloads] (LOW): The skill recommends using Icons8 (img.icons8.com) for asset management. While this is an external domain, it is a well-known service for UI icons and is used here for its intended purpose of providing tab bar icons.
  • [Indirect Prompt Injection] (LOW): The AI Model Invocation section contains a code template that passes raw userInput to an LLM.
  • Ingestion points: userInput variable in the model.streamText example.
  • Boundary markers: Absent; the example shows direct interpolation into the message array.
  • Capability inventory: The script environment has access to CloudBase resources and network-based file downloads.
  • Sanitization: None present in the example code.
  • Note: This is a standard implementation surface for AI features and is categorized as low risk in the context of a development template.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:42 PM