aws-wechat-article-writing

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls user-configured external LLM endpoints at runtime (writing_model.base_url, e.g. https://api.deepseek.com or https://api.openai.com) and POSTs prompts plus the WRITING_MODEL_API_KEY to them—these remote endpoints execute model code and are a required dependency for normal operation, creating a clear runtime external dependency risk.