cold-email-personalizer

Cold Email Personalized Mass Sending

Automatically generate highly personalized cold emails based on lead information, including compelling subject lines and targeted email bodies.

Trigger Conditions

Start this skill when the user says the following:

"Generate cold email"
"Personalized email"
"cold email"
"Help me write a prospecting letter"
"Mass email"
"outreach email"

Workflow

┌─────────────────────┐
│  读取潜在客户数据     │
│  (CSV/JSON 文件)     │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  筛选未处理的客户     │
│  (status != "sent") │
└──────────┬──────────┘
           │
           ▼
    ┌──────┴──────┐
    │  遍历每位客户  │
    └──────┬──────┘
           │
           ▼
┌─────────────────────┐
│  收集客户背景信息     │
│  (公司/职位/行业)    │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  AI 生成个性化邮件   │
│  - 主题行            │
│  - 邮件正文          │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  输出结构化结果       │
│  保存到输出文件       │
└─────────────────────┘

Data Format

Input file format (leads.csv or leads.json)

CSV Format:

name,email,company,position,industry,status
张三,zhang@example.com,科技有限公司,产品经理,SaaS,pending
李四,li@example.com,电商平台,运营总监,电商,pending

JSON Format:

{
  "leads": [
    {
      "name": "张三",
      "email": "zhang@example.com",
      "company": "科技有限公司",
      "position": "产品经理",
      "industry": "SaaS",
      "linkedin": "https://linkedin.com/in/zhangsan",
      "status": "pending"
    }
  ]
}

Output Format

{
  "emails": [
    {
      "to": "zhang@example.com",
      "name": "张三",
      "subject": "关于提升产品团队效率的一点想法",
      "body": "张三你好，\n\n我注意到科技有限公司最近在扩张产品线...",
      "generated_at": "2025-12-29T10:30:00Z"
    }
  ]
}

Execution Steps

Step 1: Prepare customer data

Check whether the user has provided a data file path:

If a file path is provided, read the file
If not provided, ask the user:
- Provide a CSV/JSON file path
- Or directly enter customer information

Step 2: Read and parse data

Use the Read tool to read the customer data file:

读取文件 → 解析 CSV/JSON → 提取客户列表

Validate required fields:

name (required)
email (required)
company (optional but recommended)
position (optional)
industry (optional)

Step 3: Filter customers to be processed

Filter conditions:

status is not "sent"
status is not "replied"
or status is empty/pending

Step 4: Collect customer background (optional enhancement)

If the customer information is not detailed enough, you may optionally supplement the information:

Use WebSearch to search:

"[客户姓名] [公司名] LinkedIn"
"[公司名] 最新动态 融资"

Extract:

Company recent achievements/news
Customer's career highlights
Industry pain points

Step 5: Generate personalized emails

For each customer, use the following prompt to generate the email:

Email generation prompt:

你是一位专业的商务开发专家，擅长撰写高转化率的冷邮件。

请根据以下潜在客户信息，生成一封个性化的冷邮件：

【客户信息】
- 姓名: {name}
- 邮箱: {email}
- 公司: {company}
- 职位: {position}
- 行业: {industry}
- 其他信息: {additional_info}

【我们的产品/服务】
{product_description}

【邮件要求】
1. 主题行要简短有力，引起好奇心，避免垃圾邮件关键词
2. 开头要个性化，体现你了解对方
3. 痛点切入要精准，基于对方行业/职位
4. 价值主张要清晰，3秒内让对方明白能获得什么
5. CTA（行动号召）要具体，比如"15分钟快速交流"
6. 整体不超过150字（中文）/ 100词（英文）
7. 语气专业但不死板，像朋友推荐而非推销

【输出格式】
请严格按照以下 JSON 格式输出：
{
  "subject": "邮件主题行",
  "body": "邮件正文内容"
}

Step 6: Organize output results

Summarize all generated emails into structured output:

{
  "generated_at": "2025-12-29T10:30:00Z",
  "total_leads": 10,
  "processed": 8,
  "skipped": 2,
  "emails": [
    {
      "to": "zhang@example.com",
      "name": "张三",
      "subject": "...",
      "body": "...",
      "personalization_points": ["提到了公司最近融资", "针对产品经理痛点"]
    }
  ]
}

Step 7: Save results

Save the output file to:~/.claude/cache/cold-emails/

File naming:emails-{日期}-{时间}.json

Also generate a preview file (Markdown format) for manual review:

# 冷邮件生成结果

**生成时间**: 2025-12-29 10:30
**处理数量**: 8/10

---

## 邮件 #1 - 张三

**收件人**: zhang@example.com
**主题**: 关于提升产品团队效率的一点想法

**正文**:
张三你好，

我注意到科技有限公司最近在扩张产品线...

---

Usage Examples

Example 1: Batch generation from a file

用户: 帮我根据 /path/to/leads.csv 生成冷邮件，我们的产品是一款 AI 客服工具

Claude:
1. 读取 leads.csv，发现 15 位潜在客户
2. 筛选出 12 位未处理客户
3. 逐一生成个性化邮件
4. 保存结果到 ~/.claude/cache/cold-emails/
5. 显示预览供用户审核

Example 2: Instant generation for a single customer

用户: 帮我给李总写一封冷邮件，他是 ABC公司 的 CEO，主要做跨境电商

Claude:
[直接根据提供的信息生成个性化邮件]

Example 3: Batch generation and export

用户: 生成冷邮件并导出为 CSV，方便导入邮件工具

Claude:
[生成邮件后，额外导出 CSV 格式]

Email Quality Checklist

After generating emails, automatically check:

Subject line length < 50 characters
Subject line contains no spam words (free, promotion, limited time, etc.)
Body has personalized elements (name, company, industry)
Has a clear value proposition
Has a clear CTA
Body length is moderate (100-200 words)

Best Practices

Improve open rate

Start subject lines with lowercase (looks more like a personal email)
Include specific numbers or time-sensitive words
A/B testing: generate 2-3 subject line candidates for the same customer

Improve reply rate

Show at the beginning that you have done research (mention the company's recent updates)
Pain points should be described specifically, not in vague generalities
The CTA threshold should be low (a 15-minute chat vs. purchasing a product)

Avoid the spam folder

Avoid all caps
Avoid too many exclamation marks
Avoid marketing buzzwords (free trial, limited-time offer, etc.)
Do not include attachments in the first email

Data storage

Output directory:~/.claude/cache/cold-emails/
JSON result:emails-{YYYYMMDD}-{HHMMSS}.json
Markdown preview:emails-{YYYYMMDD}-{HHMMSS}.md
CSV export (optional):emails-{YYYYMMDD}-{HHMMSS}.csv

Dependencies

Read: Read the customer data file
WebSearch: (Optional) Supplement customer background information
Write: Save the generated email results

Limitations

This skill generates email content and is not responsible for actual sending
It is recommended to review manually before sending
For large-volume sending, please comply with your email service provider's sending limits
Please ensure compliance with local anti-spam regulations (such as GDPR, CAN-SPAM)