nsfc-humanization

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests arbitrary user-provided text as its primary input (see SKILL.md "执行流程 1. 读取用户提供的文本" and README examples asking users to paste their text), so untrusted user-generated or third‑party content would be read and interpreted as part of the workflow and could carry embedded instructions despite the stated guardrails.