告别复制粘贴：将 OpenAI Prompt 转化为 Claude Code 智能技能 (实战指南)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19

---
name: <技能名称，如 generate-prd>
description: <一句话描述，如：为新功能撰写产品需求文档 (PRD)>
---

# <技能标题>

## When to use this skill
- <描述触发场景1，例如：当用户要求写 PRD 时>
- <描述触发场景2，例如：当用户提供了功能想法但需要文档化时>

## Instructions
<这里粘贴 OpenAI 的 Prompt 内容>
- Context: ...
- Task: ...
- Output Format: ...

## Input Variables
- <变量1>: <描述>

1

Based on this feature idea and customer need, write a first-draft PRD. Include user story, problem statement, solution overview, acceptance criteria, and success metrics. [Insert context or problem]

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

---
name: product-draft-prd
description: Draft a comprehensive Product Requirements Document (PRD) from a feature idea.
---

# Product Requirements Document Generator

## When to use this skill
- When the user asks to write a PRD or spec for a feature.
- When the user provides a rough feature idea and needs a structured document.
- When asking to "flesh out" a product concept.

## Instructions
You are an expert Product Manager. Your goal is to draft a PRD based on the user's input.

1.  **Analyze the Request**: Identify the core feature, target audience, and business goal from the user's prompt.
2.  **Draft the Content**: Generate a response following this strict structure:
    *   **Problem Statement**: What are we solving?
    *   **User Stories**: As a [user], I want to [action], so that [benefit].
    *   **Acceptance Criteria**: Gherkin syntax (Given/When/Then) or bullet points.
    *   **Technical Constraints**: Potential limitations or requirements (ask if unknown).
    *   **Success Metrics**: How will we measure success?

3.  **Refinement**: If the user's input is too vague, ask clarifying questions before generating the full doc.

1

帮我写一份关于‘AI 自动剪辑短视频’功能的 PRD

1

Explain this code snippet to a junior engineer, point out potential bugs, and suggest a refactor for better performance.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17

def process_data(l):
    # 保存去重后的用户
    res = []
    for u in l:
        # 糟糕的去重逻辑：每次循环都遍历一遍 res，性能灾难
        if u['id'] not in [x['id'] for x in res]:
            res.append(u)

    total = 0
    count = 0
    for u in res:
        if u['active'] == True:
            total += u['age']
            count += 1

    # 致命 Bug：如果 count 为 0，这里会报错 (ZeroDivisionError)
    return total / count