The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to install the uview-pro package via npm (npm install uview-pro) and references the official uView Pro plugin on the DCloud marketplace (ID: 24633). These are legitimate resources within the UniApp ecosystem.
[DATA_EXFILTRATION]: Examples in examples/integration/uniappx-api.md and examples/platform-specific/miniprogram.md demonstrate using uni.request to fetch data from external APIs (e.g., https://api.example.com/data). While these are standard developmental practices, they represent network operations to non-whitelisted domains.
[PROMPT_INJECTION]: (Category 8
Indirect) The skill demonstrates patterns for ingesting untrusted data which could serve as an attack surface for indirect prompt injection.
Ingestion points: Data enters the application context through uni.request in examples/integration/uniappx-api.md and through URL parameters processed via onLoad in examples/integration/navigation.md.
Boundary markers: None are present in the provided code templates to differentiate between data and instructions.
Capability inventory: The skill patterns utilize network (uni.request), local storage (uni.setStorage), and navigation (uni.navigateTo) APIs across various files.
Sanitization: The examples do not include logic for sanitizing or validating ingested external data before it is rendered or used in application logic.

uniappx-uview-pro