short-drama

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly asks the agent to accept a user-pasted MCP API key (wwmcp_...), validate it, and write that key into config files via Edit tools, which requires the LLM to receive and output the secret value verbatim.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These URLs are suspicious because they include direct links to raw executable/install scripts on an unknown GitHub account (install.sh, install.ps1/raw content) plus an obscure redirector (ghfast.top) and a private domain (lingjingai.cn) — indicators commonly used to distribute malware, while the w3.org/schemas and pandoc/markdowntoword links are benign references that do not offset the high-risk items.