content-engine

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

• Third-party content exposure detected (medium risk: 0.30). Skill 以 modes/capture.md 的 Capture（“收真料”）作为唯一真证据入口，运行时会读取用户/业务方提供的客户话语并把原话与出处标记写入 Content-Engine-Brand-Memory.md，属于可被用户提交的自由文本证据输入路径。

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill explicitly instructs modifying an existing installed skill file in the user's home (~/.claude/skills/-command-base/SKILL.md) during installation, which changes the machine/user state and could alter other installed software.