article-batch-illustration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes an example that embeds an API key as a command-line argument (--api-key "GEMINI_API_KEY") and references storing keys in files, which instructs copying secret values into generated commands/outputs (high exfiltration risk).

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md Phase 1 “读取文章”直接摄取用户提供文章的自由文本（按 H2/H3 拆分并提取核心观点/关键概念），随后将这些内容写入 Gemini/图像生成提示词作为 LLM 输入，因此可能被用户投喂间接提示注入。