ai-short-drama

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill instructs users to fetch-and-execute a remote installer via "curl | bash" (jimeng.jianying.com/cli) and to run third‑party CLI commands from scripts, which is a high-risk supply‑chain / remote‑execution pattern that can enable backdoors or arbitrary code installation.

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md 的运行流程包含“市场情报扫描”阶段会对网页内容进行 WebFetch/真读爆款文章（外部网页抓取属于 public web content fetched at runtime），这些抓取到的自由文本会被用于后续生成/写入并进入 LLM 上下文的概率较高。

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

• Potentially malicious external URL detected (high risk: 1.00). The skill requires the "dreamina" CLI at runtime and explicitly instructs installing it via "curl -fsSL https://jimeng.jianying.com/cli | bash", which fetches and executes remote code from https://jimeng.jianying.com/cli as a required dependency.