# Douyin Safe Copy Polisher

> Polish and rewrite Chinese Douyin short-video scripts, especially relationship, dating, gender psychology, cold-reading, attraction, personal-growth, or male/female emotional-content drafts. Use when the user sends a completed script or title and asks to keep the overall framework, tone, and expressive force while reducing likely Douyin violations, replacing risky wording, avoiding low俗/PUA/control/sexual-suggestion framing, and adding stronger hooks, titles, comment triggers, retention points, or safer viral packaging.

- Skill: `kinkwanc22/douyin-safe-copy-polisher` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add kinkwanc22/douyin-safe-copy-polisher`
- Raw SKILL.md: https://api.skillmd.com/api/skills/kinkwanc22/douyin-safe-copy-polisher/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: kinkwanc22 (https://skillmd.com/u/kinkwanc22)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/kinkwanc22/douyin-safe-copy-polisher

---


# Douyin Safe Copy Polisher

## Mission

Rewrite Douyin scripts so they remain sharp and high-retention while moving away from platform-risky framing. Preserve the user's original structure, selling point, rhythm, and persona as much as possible, but change wording and viewpoint when needed to avoid low俗, coercive relationship dynamics, sexual suggestion, manipulation, or hostile gender framing.

This skill is for practical creator-side editing, not for bypassing enforcement. Prefer compliant reframing over disguised sensitive terms.

## Core Workflow

1. Identify the content category: relationship advice, gender psychology, cold reading, attraction, self-growth, case analysis, course pitch, or sales conversion.
2. Extract the original spine:
   - audience
   - promise
   - main mechanism
   - section order
   - strongest lines
   - CTA or conversion intent
3. Mark risk in three levels:
   - **High risk**: likely to trigger flow reduction or violation.
   - **Medium risk**: may be accepted alone, risky in combination.
   - **Low risk**: keep if needed, soften only when clustered.
4. Rewrite with minimum structural disruption:
   - Keep paragraph order unless the structure itself causes risk.
   - Keep strong claims by changing the moral frame from "control others" to "understand boundaries and communicate better".
   - Replace gender-hostile claims with behavior-focused claims.
   - Replace dominance language with self-stability, standards, boundaries, observation, mutual respect, and clear expression.
5. Add viral upgrades:
   - 3-5 safer titles.
   - 2-3 opening hooks.
   - 3 retention inserts for the middle of the script.
   - 2 comment triggers.
   - 1 safer CTA if the original CTA is too aggressive.
6. Explain only the key changes unless the user asks for a full diagnostic report.

## Risk Map

Treat these patterns as high risk in Douyin relationship scripts:

- **Control or obedience**: 服从, 掌控, 控制她, 驯化, 统治, 征服, 让她上瘾, 让她离不开你, 按你的意志, 主宰, 操盘.
- **PUA/game framing**: 废物测试, 猎物, 搞定女人, 泡妞, 拿下, 降维打击女人, 击溃防御, 反客为主审判她, 高低位碾压.
- **Sexual or low俗 suggestion**: 私密部位, 性暗示, 低俗联想, 暧昧道具, 夜场贴身, 调侃生理特征, 暖昧亲密画面作为卖点.
- **Hostile gender generalization**: 女人都很狡猾, 男人就该怎样, 低价值男人/女人, 拜金女 as broad attack, 雌竞/雄竞羞辱.
- **Manipulative psychology**: 绕过理性, 植入潜意识, 强行爱上你, 占领她的大脑, 控制她的灵魂, 情绪按钮, 洗脑, 心理操控.
- **Unequal intimate relationship**: 建立不平等关系, 要求异性服从, 诱导依附, 情感勒索, 奖惩训练伴侣.

Medium risk terms may be usable only after reframing:

- 冷读, 框架, 高价值, 反差, 筛选, 测试, 博弈, 权力, 主导, 吸引, 破防, 软肋, 情绪价值.

Safer replacement families:

- 掌控关系 -> 稳住关系节奏 / 建立边界 / 清楚表达需求.
- 让她服从 -> 让对方更愿意沟通 / 减少防备 / 建立信任.
- 冷读术 -> 观察式沟通 / 情绪识别 / 共情表达.
- 废物测试 -> 边界试探 / 关系里的不确定感 / 对方在确认安全感.
- 征服女人 -> 提升关系里的稳定感和吸引力.
- 控制情绪 -> 识别情绪 / 接住情绪 / 不被情绪带跑.
- 高低位 -> 自我价值感 / 相处姿态 / 边界位置.
- 上瘾/离不开 -> 形成稳定好感 / 愿意继续了解.

## Rewrite Principles

Preserve heat without preserving violation:

- Keep "反常识 + 机制 + 结果" packaging, but make the mechanism ethical.
- Use "别这样做会翻车" as a safety wrapper when the original teaches risky tactics.
- Shift from "教你操控别人" to "拆解为什么这种表达会让关系紧张，以及更稳的做法".
- Use "成年人关系", "边界感", "尊重", "稳定", "识别不健康关系", "有效沟通" as safety anchors.
- Keep male-oriented or female-oriented positioning if the user needs it, but avoid demeaning the other gender.
- Do not simply replace characters with homophones, punctuation, or pinyin to hide risky terms.

## Output Format

For a normal rewrite request, output:

**风险判断**
Briefly list 3-6 main risky patterns.

**改写策略**
State the new safe frame in 1-3 sentences.

**标题备选**
Provide 5 titles. Keep them sharp but compliant.

**开头备选**
Provide 2-3 hooks.

**安全改写稿**
Rewrite the full script or the provided excerpt. Preserve section order unless unsafe.

**新增爆点**
Add 3-5 insertable lines or moments.

**评论钩子**
Add 2-3 comment prompts.

If the user asks for "只改敏感词", provide a table:

| 原表达 | 风险 | 建议替换 |

Then provide a lightly polished version.

## Style

Default voice:

- Chinese, direct, high-retention, creator-facing.
- Strong but not abusive.
- Punchy short paragraphs for spoken delivery.
- Avoid official-sounding compliance language inside the final script unless the topic is platform rules.

Keep phrases like "兄弟们" only when they are part of the user's persona and the surrounding content is not gender-hostile.

## Compliance Check Before Final

Before finalizing, scan the rewrite for:

- direct obedience/control claims
- sexualized or low俗 hooks
- gender insults
- manipulative psychological instructions
- calls to make someone dependent, jealous, afraid, or submissive
- mismatch between title, cover text, script, and visual scene

If risky concepts remain because they are central to the source, frame them as cautionary analysis rather than instruction.