SuperAPI AI Directory

title-designer

Writing & research

View repository

Updated 10-15 21:11

About this skill

Title Designer

Instructions

Core goal: Design a title for the article that can grab readers. The title is the article's "front door", determining whether readers will click.

1. Input confirmation

Before designing the title, confirm the following information:

📋 标题设计输入确认:

• 文章主题:[从前置流程获取]
• 核心观点:[从前置流程获取]
• 目标读者:[从前置流程获取]
• 写作风格:[从前置流程获取,如:九边风/墨水怪风/无特定风格]

Is the information complete? (Yes/Supplement)


如果信息不完整,询问用户补充。

---

### 2. 设计3个候选标题

根据输入信息,设计3种不同风格的标题:

📝 Please choose an article title:

A.【Catchy】____________________ Strategy: Use suspense, rhetorical questions, or surprising data to attract clicks Example: "35-year-old programmer crisis? Wake up, you've been harvested"

B.【Informative】____________________ Strategy: Directly state the main point of the article so readers know what they will read Example: "Stack Overflow data tells you: what is this 35-year-old crisis thing"

C.【Emotional】____________________ Strategy: Touch emotions, trigger resonance or curiosity Example: "Those people selling 35-year-old anxiety earn in one year what you earn in several lifetimes"

Please choose A/B/C, or tell me the title direction you want:


---

### 3. 标题设计规则

**禁止使用的AI套路标题:**
- ❌ "深度解读XXX"
- ❌ "全面分析XXX"
- ❌ "你不知道的XXX"
- ❌ "震惊!XXX"
- ❌ "XXX的10个真相"
- ❌ "关于XXX,你需要知道的一切"

**推荐技巧:**
- ✅ 用具体数字增强可信度("1000人公司"比"中小企业"更具体)
- ✅ 用反问制造悬念("老板装监控的那天,公司就开始死了?")
- ✅ 用对比制造冲突("你以为是韭菜,其实是镰刀")
- ✅ 用熟悉的场景引发共鸣("那张总监名片,不过是租来的座位")

**长度建议:**
- 中文标题:15-25字
- 太短没信息量,太长读不完

**风格适配:**
- 如果使用了风格文件(如九边风、墨水怪风),标题也要符合该风格的语气
- 九边风:偏务实、接地气,带点无奈
- 墨水怪风:偏犀利、反常识,敢骂人

---

### 4. 用户选择后

用户选择标题后,输出确认:

✅ Title confirmed: "[Title chosen by the user]"

This title will be used for: • Top of the article body • File naming: articles/[project name]/draft_v1.md

Continue to the next stage? (Yes/Modify title)


---

### 5. 标题优化(可选)

如果用户对3个候选都不满意,可以:

1. **追问方向**:"您希望标题更偏向哪个方向?(更犀利/更温和/更具体/更抽象)"
2. **微调现有**:基于用户反馈,微调某个候选标题
3. **用户自拟**:用户提供标题,AI帮助优化

---

## 示例 (Examples)

**输入:**
- 主题:中小企业从大厂招人后的认知落差
- 观点:中小企业觉得招到名校生是"升级",但求职者眼里这还是"中小企业"
- 读者:职场人/HR/创业者
- 风格:九边风

**输出:**

📝 Please choose an article title:

A.【Catchy】A 1000-person company starts recruiting from prestigious schools, but this is not a good thing Strategy: Counterintuitive, create cognitive conflict

B.【Informative】The "prestigious school illusion" of small and medium-sized enterprises: you think you've upgraded, but actually you've just switched leeks Strategy: Directly point out the theme, use metaphor to strengthen the memory point

C.【Emotional】That post-95er who jumped from ByteDance looks at you like looking at a vegetable market Strategy: Specific scene + emotional impact, create a sense of immersion

Please choose A/B/C, or tell me the title direction you want:


---

## 最佳实践 (Best Practices)

- **3选1是底线**:必须给出3个明显不同风格的候选,让用户有选择空间
- **禁止自动选择**:必须等用户确认,不能代替用户决定标题
- **风格优先**:如果有风格文件,标题必须符合该风格的语气
- **具体胜过抽象**:"1000人公司"比"中小企业"更有画面感

## 版本记录 (Version History)
- v2.0.0 (2026-01-03): 新增爆款标题公式库(5种类型)、标题自检清单、平台适配建议。
- v1.0.0 (2025-12-28): 初始版本,独立标题设计Skill。

---

## 爆款标题公式库 ⭐⭐⭐⭐⭐(核心能力!)

**除了基础的吸引型/信息型/情绪型,必须至少使用1个爆款公式:**

### 公式1:数字具体化型
**模板:** `「数字 + 身份 + 痛点/结果」`

| 示例 | 触发机制 |
|------|---------|
| "月薪3万的人,都戒掉了这3个习惯" | 具体数字增强可信度 |
| "我面试了100个候选人,只录用了这3种人" | 大数字制造权威感 |
| "35岁程序员,年薪从80万跌到15万" | 数字对比制造冲击 |

### 公式2:对立冲突型
**模板:** `「你以为是A,其实是B」`

| 示例 | 触发机制 |
|------|---------|
| "你以为是努力,其实是自毁" | 认知反转,打破舒适区 |
| "你以为是996,其实是在为35岁裁员做准备" | 揭示隐藏真相 |

### 公式3:身份认同型
**模板:** `「XX的人」+ 状态/结局`

| 示例 | 触发机制 |
|------|---------|
| "30岁还在写代码的人,后来都怎样了" | 引发身份焦虑/好奇 |
| "从大厂裸辞的人,3年后都在干什么" | 替代人生路径探索 |

### 公式4:悬念反转型
**模板:** `「动作 + 结果反转」`

| 示例 | 触发机制 |
|------|---------|
| "我花3年考下注册会计师证,然后失业了" | 努力与结果反差 |
| "当上总监后,我反而想辞职了" | 成功后的反转 |

### 公式5:情绪宣泄型
**模板:** `「直接说出读者想骂的话」`

| 示例 | 触发机制 |
|------|---------|
| "职场最恶心的那句话,不是996,而是…" | 替读者发泄 |
| "老板,你装的摄像头,我们都知道" | 揭露潜规则 |

---

## 标题自检清单(发给用户前必须检查!)

📋 Title self-check list:

□ 1.【3-second rule】When users scroll to this title, will they stop within 3 seconds? □ 2.【Concretization】Have specific numbers/scenes been used instead of abstract concepts? □ 3.【Identity alignment】When target readers see the title, will they feel "this is talking about me"? □ 4.【Suspense/Conflict】Does the title create curiosity or cognitive conflict? □ 5.【Taboo word check】Have AI cliché words such as "in-depth interpretation" or "comprehensive analysis" been used?

✅ All 5 items passed → Can be used ❌ Fewer than 3 items passed → Needs redesign


---

## 平台适配建议

| 平台 | 标题特点 | 建议公式 |
|------|---------|---------|
| 公众号 | 情绪驱动 | 情绪型/身份认同型 |
| 知乎 | 专业感 | 数字具体化型 |
| 今日头条 | 悬念有效 | 悬念反转型 |
| 小红书 | 短标题、"我"视角 | 身份认同型 |