🎭 Domain Expert Prompts | 打造领域专家提示词

🎯 Learning Objective | 学习目标：Master the art of crafting prompts that turn AI into domain experts | 掌握让AI变身领域专家的提示词技艺

---

🌟 What is Domain Expert Prompting? | 什么是领域专家提示词？

Imagine you could hire any expert instantly: 想象你可以即时聘请任何专家：

• 🏥 Medical Consultant | 医疗顾问 - For health-related questions | 解答健康问题
• 💼 Business Strategist | 商业策略师 - For career decisions | 职业决策
• 💻 Senior Developer | 资深开发者 - For coding challenges | 编程挑战
• 📝 Editor | 编辑专家 - For writing improvement | 写作提升

With the right prompt, AI becomes that expert! 用对了提示词，AI就变成那个专家！

---

📋 The Expert Prompt Framework | 专家提示词框架

CO-STAR Method | CO-STAR方法

Element	Description	Example
元素	描述	示例
Context	Background information	"I'm a QA engineer..."
C上下文	背景信息	"我是一名测试工程师..."
Objective	What you want to achieve	"Help me write test cases"
O目标	你想达成什么	"帮我编写测试用例"
Style	How AI should respond	"Be detailed and structured"
S风格	AI应如何回应	"详细且结构化"
Tone	The voice/personality	"Professional but friendly"
T语气	语气/人设	"专业但友好"
Audience	Who will read the output	"Junior developers"
A受众	谁会阅读输出	"初级开发者"
Response	Output format	"Markdown with code blocks"
R响应	输出格式	"带代码块的Markdown"

---

🎯 Practical Examples | 实战示例

Example 1: Programming Expert | 示例1：编程专家

<role>
You are a senior Python developer with 10+ years of experience.
你是一位拥有10年以上经验的资深Python开发者。
Expertise: Clean code, design patterns, performance optimization.
专长：整洁代码、设计模式、性能优化。
</role>

<constraints>
- Always follow PEP 8 standards
- 始终遵循PEP 8标准
- Add type hints to all functions
- 为所有函数添加类型提示
- Include docstrings and comments
- 包含文档字符串和注释
</constraints>

<output_format>
- Brief explanation first
- 先简要说明
- Then complete code
- 然后完整代码
- Finally, usage examples
- 最后使用示例
</output_format>

Example 2: Learning Tutor | 示例2：学习导师

<role>
You are a patient and experienced teacher.
你是一位耐心且经验丰富的老师。
Skilled at explaining complex concepts simply.
擅长用简单的方式解释复杂概念。
</role>

<teaching_style>
- Use analogies from daily life
- 使用生活中的类比
- Break down into small steps
- 分解为小步骤
- Check understanding frequently
- 经常检查理解程度
- Encourage questions
- 鼓励提问
</teaching_style>

<audience>
Student with basic knowledge, learning [TOPIC].
有基础知识的学生，正在学习[主题]。
</audience>

Example 3: Research Assistant | 示例3：研究助手

<role>
You are an academic research assistant.
你是一位学术研究助手。
Expert in literature review and synthesis.
专长于文献综述和综合分析。
</role>

<task>
When given a research topic:
当给定一个研究主题时：
1. Identify key concepts and terms
   识别关键概念和术语
2. Suggest search strategies
   建议搜索策略
3. Summarize findings objectively
   客观总结发现
4. Point out knowledge gaps
   指出知识空白
</task>

<output_rules>
- Cite sources when possible
- 尽可能引用来源
- Distinguish facts from opinions
- 区分事实和观点
- Acknowledge limitations
- 承认局限性
</output_rules>

---

🏗️ Building Your Prompt Library | 构建你的提示词库

Step 1: Start with Templates | 从模板开始

📁 My Prompt Library
├── 💻 Programming/
│   ├── code_reviewer.md
│   ├── debugger.md
│   └── architect.md
├── 📝 Writing/
│   ├── editor.md
│   ├── translator.md
│   └── summarizer.md
├── 🎓 Learning/
│   ├── tutor.md
│   ├── explainer.md
│   └── quiz_master.md
└── 🔬 Research/
    ├── literature_reviewer.md
    ├── data_analyst.md
    └── brainstormer.md

Step 2: Iterate and Improve | 迭代优化

Initial Prompt → Test → Analyze Output → Refine → Test Again
初始提示词   → 测试 → 分析输出     → 优化  → 再次测试

Key Questions for Improvement | 优化关键问题： - Did AI understand the context? | AI理解了上下文吗？ - Was the output format correct? | 输出格式正确吗？ - Did it miss any important aspects? | 是否遗漏了重要方面？ - Was it too verbose or too brief? | 是否太冗长或太简短？

---

🔥 Advanced Techniques | 进阶技巧

1. Chain of Thought (CoT) | 思维链

Force AI to think step-by-step: 强制AI逐步思考：

<thinking>
Before answering, please:
回答前，请：
1. Analyze the problem components
   分析问题组成
2. Consider multiple approaches
   考虑多种方法
3. Evaluate pros and cons
   评估利弊
4. Then provide your recommendation
   然后提供你的建议
</thinking>

2. Few-Shot Learning | 少样本学习

Provide examples to guide output: 提供示例来引导输出：

Example 1:
Input: "def calc(a,b): return a+b"
Output: "def calculate_sum(num1: int, num2: int) -> int:
    '''Calculate the sum of two numbers.'''
    return num1 + num2"

Example 2:
Input: "def f(x): return x*x"
Output: "def calculate_square(number: float) -> float:
    '''Calculate the square of a number.'''
    return number * number"

Now apply this to: [YOUR CODE]
现在应用于：[你的代码]

3. Self-Consistency | 自我一致性

Ask AI to verify its own answers: 让AI验证自己的答案：

<verification>
After providing your answer:
提供答案后：
1. Re-read the original question
   重读原始问题
2. Check if your answer fully addresses it
   检查答案是否完全解决了问题
3. Identify any assumptions made
   识别做出的任何假设
4. Rate your confidence (1-10)
   给出你的信心评分（1-10）
</verification>

---

📚 Real-World Prompt Collection | 实战提示词收藏

For Code Review | 代码审查

You are a strict but helpful code reviewer.
Review this code for:
- Security vulnerabilities
- Performance issues
- Code style violations
- Potential bugs
- Improvement suggestions

For each issue found, explain WHY it's a problem 
and provide a concrete fix.

For Document Summarization | 文档总结

Summarize the following document using this structure:
1. **Key Points** (3-5 bullet points)
2. **Main Arguments** (brief paragraph)
3. **Conclusions** (1-2 sentences)
4. **Questions Raised** (for further exploration)

Keep the summary under 300 words.
Preserve technical accuracy.

For Learning New Topics | 学习新主题

Explain [TOPIC] to me as if I'm a smart 12-year-old.
- Use simple analogies
- Avoid jargon (or explain it)
- Give real-world examples
- End with a small quiz to check my understanding

After your explanation, ask me if I have questions.

---

⏱️ Practice Exercises | 练习任务

1. Create Your First Expert | 创建你的第一个专家
2. Pick a domain you need help with
3. Write a CO-STAR prompt
4. Test it with 3 different questions

5. Iterate to improve

6. Build a Prompt Library | 构建提示词库

7. Create 5 prompts for your daily work
8. Organize them in a system

9. Share with colleagues for feedback

10. Advanced Challenge | 进阶挑战

11. Combine multiple techniques (CoT + Few-shot)
12. Create a self-improving prompt that learns from feedback

---

💡 Pro Tip | 小贴士：The best prompts are like good documentation - clear, specific, and with examples. Treat prompt writing as a skill to be practiced!

最好的提示词就像好的文档——清晰、具体、有示例。把提示词写作当作一门需要练习的技能！

---

📖 Recommended Resources | 推荐资源

• System Prompts Collection - Learn from top AI products | 从顶级AI产品学习
• Awesome Prompts - Community-curated prompts | 社区精选提示词
• OpenAI Cookbook - Official best practices | 官方最佳实践