wechat-ai-radar

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The code automates capturing WeChat Moments (screenshots and extracted text) and intentionally sends collected social content to a third‑party AI API (https://api.qingyuntop.top) — an explicit data‑exfiltration / privacy‑invasive behavior.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md:方式一/方式三/方式四会运行生成简报脚本（如 scripts/wechat_radar_briefing.py），该脚本从外部来源的 extracted_*.json 读取朋友圈“content/author_name”并拼入 LLM prompt（call_ai_analysis: content_text → prompt → httpx.post messages.user.content）。