The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation in references/framework.md describes the installation of miniprogram-simulate, which is the standard official utility for unit testing WeChat Mini Programs, from the npm registry.
[INDIRECT_PROMPT_INJECTION]: This skill provides instructions for a development framework that processes user-controlled files (WXML, JS). This constitutes an attack surface for indirect prompt injection if the agent uses these instructions to generate or analyze code from untrusted sources without proper boundaries.
Ingestion points: Mini Program project files including SKILL.md examples and the structure defined in references/getting_started.md.
Boundary markers: None explicitly defined in the provided documentation to differentiate between developer instructions and potentially untrusted code content.
Capability inventory: The skill documents powerful APIs for network communication (wx.request), file system access, and cloud function execution as detailed in references/api.md and references/cloud.md.
Sanitization: Not applicable for this documentation-only skill, which serves as a technical reference.

wechat-miniprogram