qqmusic

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt asks the agent to request the user's API key and to write an export line containing that key into shell profile files (and thus emit the key in commands/outputs), which requires the LLM to handle and output the secret verbatim.

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

• Third-party content exposure detected (high risk: 0.75). 该 skill 的运行时会在用户明确“分析/解读/画像”时调用 /assistant/ai_interpretation，并将用户原话（外部用户输入的自由文本）作为 params.query 进入 LLM 上下文，属于 outsider-authored free text 的间接提示注入风险。