结尾
AI 改变的是取得初稿的速度，不是作出专业判断的能力。工具走到边界，留下的仍是判断与责任。
AI 最危险的时候，是它看起来还能使用的时候。
/de-ai-polish
@6113db6检测并修订中文文章正文中的模板腔、姿态腔、隐藏排比、分析型例证同构、重复语义骨架、假排名、固定节拍和其他 AI 化表达，同时保护作者事实、既有标题层级、必要功能性列举、法律术语、Markdown 图片与既有声音。文章完成初稿后、用户要求“去 AI 腔/更像真人/按作者样本改写/检查排比或口吻”时必须使用；无作者样本时只做最小清理，不得凭空注入第一人称、比喻、短句金句或经历，也不得借去 AI 之名重做文章结构。
Ask your Agent
Use this Skill: https://skilld.dev/gh/cat-xierluo/legal-skills/de-ai-polish
This session only. Nothing lands on disk.
≈52 tokens on demand. Your agent reads this file only when SKILL.md points to it.
Source: SKILL.md on GitHub
Third-party checks
1 warning15d4 checks · Risk SAFE
- Gen Agent Trust Hub15d
The de-ai-polish skill is a sophisticated NLP toolkit designed to naturalize AI-generated Chinese text. It utilizes a robust framework of internal validation scripts to ensure Markdown structure preservation (images and headings) and stylistic integrity without relying on external dependencies or performing risky network operations. The implementation follows security best practices, including privacy-preserving hashing for logs and defensive coding to prevent data leakage.
- Socket15d
No alerts
- Snyk15d
Risk: LOW · No issues
- Runlayer7mo
6/6 files flagged
Provenance
Signed by skilld at 6113db6. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.
Last checked against GitHub yesterday.
Activeupdated 2 months ago
Capability
Other metadata
- metadata
{ "author": "杨卫薪律师（微信ywxlaw）", "homepage": "https://github.com/cat-xierluo/legal-skills", "version": "3.2.6" }