hkr-render

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The script comment_reply.py automatically scans public WeChat comments from published articles and processes them with an LLM using system prompt instructions, creating an indirect prompt injection vector via untrusted reader-authored text.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (medium risk: 0.30). The skill executes git clone from https://github.com/xiaohuailabs/xiaohu-wechat-format.git during installation instructions, which is an unofficial third-party repository and represents an unverified runtime external dependency source.