ai-product-skills

适用于产品/AI产品同学的工作流 | Workflow skills for AI product managers

中文

仓库结构

ai-product-skills/
├── product-methodology/        # 产品方法论 + 方案审查
│   ├── PRINCIPLES.md           # 产品原则（通用）
│   ├── SOP.md                  # 产品方案 SOP（通用）
│   ├── REVIEW.md               # UI/体验 Review 清单（通用）
│   └── review-plan-SKILL.md    # 海外版方案审查 skill
└── intern-experience/          # 实习经历沉淀（产品经理实习生专用）
    ├── cv-skill.md             # 工作记录 → 简历项目经历
    └── intern-introduction.md  # 面试口头介绍 + 模拟考核

Skills 列表

本仓库包含六个 skill，分为三类：

#### 产品方法论（product-methodology/） — 通用

一套完整的产品方法论体系，三个文档各司其职、互相引用，适用于任何产品方向：

使用方式：

• 做方案前：读 PRINCIPLES.md，用原则校准方向
• 做方案时：按 SOP.md 四步流程推进
• 做 demo 前：把 REVIEW.md 喂给 Claude Code 当生成约束
• 做完后：用 REVIEW.md 逐层自检验收

#### 实习经历沉淀（intern-experience/） — 产品经理实习生专用

包含两个互相配合的 skill：

cv-skill.md — 工作记录 → 简历项目经历

将实习期间散落各处的工作记录（飞书待办、方案文档、会议纪要、OKR 等），整理为可直接粘进简历的 bullet points。

• 从琐碎记录中提取关键事实（做了什么、方法、产出、量化结果）
• 按项目维度结构化梳理，提炼 2-3 个最能打的卖点
• 输出三个版本：一句话版 / 标准版（3-4 行） / 详细版（5-6 行）
• 中英文双版本
• 内置面试官视角质量检查：分量感、专业度、结果可信度

intern-introduction.md — 面试口头介绍 + 模拟考核

基于 cv-skill 输出的项目经历，生成面试时的口头介绍话术和高频考核问答。

• STAR 框架口头介绍（2-3 分钟，Action 占 40-50% 篇幅）
• 5-8 个模拟考核问题（需求判断、用户理解、方案设计、数据思维等）
• 每个问题配参考回答（结论先行 + 展开逻辑 + 落地举例）
• 产品思维检查清单自检

使用场景：

• 日常积累：每 1-2 周用 cv-skill 整理一次，避免最后想不起来做了什么
• 阶段性整理：实习结束前，一次性整理所有记录为简历文本
• 面试准备：用 intern-introduction 生成口头话术和模拟问答，针对目标岗位定制

#### /review-plan — 海外版专用

一个专门针对海外版产品方案的 Claude Code skill。从创新性、海外市场适配性、方案完整性三个维度逐项检查，给出具体问题和可操作的改进建议。

适合谁用：

• 需要自主审查方案、没有 mentor review 的场景
• 正在为海外用户设计功能
• 希望用结构化清单来发现方案盲区

使用流程：

1. 将产品方案从飞书（或其他文档工具）导出为 Word 文件
2. 在 Claude Code 中调用 /review-plan，提供文件路径
3. Claude 按 6 大类清单逐项检查，在对话中输出结果
4. 和 Claude 讨论、迭代
5. 讨论完毕后，Claude 生成一份 Word 审查报告

6 大检查类别：

安装

将对应 skill 文件夹复制到 Claude Code 的 skills 目录：

# 个人使用（所有项目通用）
cp -r product-methodology ~/.claude/skills/
cp -r intern-experience ~/.claude/skills/

# 仅限当前项目
cp -r product-methodology .claude/skills/
cp -r intern-experience .claude/skills/

自定义

product-methodology/review-plan-SKILL.md 中有一个"产品背景信息"区域，预留了占位符。填入你自己的产品背景信息，可以获得更有针对性的审查：

• 产品定位
• 目标市场和用户画像
• 功能规划清单
• 海外版和国内版的主要差异
• 运营侧的要求和节奏
• 技术/合规约束

English

Repository Structure

ai-product-skills/
├── product-methodology/        # Product methodology + plan review
│   ├── PRINCIPLES.md           # Product principles (universal)
│   ├── SOP.md                  # Product plan SOP (universal)
│   ├── REVIEW.md               # UI/UX review checklist (universal)
│   └── review-plan-SKILL.md    # Overseas plan review skill
└── intern-experience/          # Intern experience (for PM interns)
    ├── cv-skill.md             # Work records → resume bullet points
    └── intern-introduction.md  # Interview oral intro + mock Q&A

Skills

This repo contains six skills in three categories:

#### Product Methodology (product-methodology/) — Universal

A complete product methodology system — three docs, each with a clear role, cross-referencing each other. Applicable to any product direction:

How to use:

• Before planning: Read PRINCIPLES.md to calibrate direction
• While planning: Follow the four-step process in SOP.md
• Before building a demo: Feed REVIEW.md to Claude Code as generation constraints
• After building: Self-review against REVIEW.md layer by layer

#### Intern Experience (intern-experience/) — For PM Interns

Two complementary skills:

cv-skill.md — Work Records → Resume Bullet Points

Converts scattered internship work records (task lists, docs, meeting notes, OKRs) into resume-ready bullet points.

• Extracts key facts from messy records (what you did, methods, deliverables, metrics)
• Organizes by project, distills 2-3 strongest selling points per project
• Outputs three versions: one-liner / standard (3-4 bullets) / detailed (5-6 bullets)
• Chinese + English dual output
• Built-in quality check from an interviewer's perspective: impact, professionalism, credibility

intern-introduction.md — Interview Oral Intro + Mock Q&A

Generates interview oral presentation scripts and mock Q&A based on cv-skill output.

• STAR framework oral introduction (2-3 min, Action at 40-50% of content)
• 5-8 mock interview questions per project (requirements judgment, user understanding, solution design, data thinking, etc.)
• Reference answers for each question (conclusion first + logic + concrete examples)
• Product thinking self-check checklist

Use cases:

• Regular accumulation: Use cv-skill every 1-2 weeks so nothing is forgotten
• End-of-internship summary: Batch process all records into resume text
• Interview prep: Use intern-introduction to generate oral scripts and mock Q&A, customize for target roles

#### /review-plan — Overseas-specific

A Claude Code skill specifically designed for reviewing overseas product plans. It checks your plan from three dimensions: innovation, overseas market fit, and completeness, then provides specific issues and actionable suggestions.

Who is this for:

• PMs who need to self-review plans without a mentor
• Anyone designing features for international users
• Those who want a structured checklist to catch blind spots

How it works:

1. Export your product plan from Feishu (or any doc tool) as a Word file
2. Call /review-plan in Claude Code and provide the file path
3. Claude checks the plan against a 6-category checklist and outputs results in chat
4. Discuss and iterate together
5. When done, Claude generates a Word review report

6 review categories:

Install

Copy the skill folder to your Claude Code skills directory:

# Personal use (all projects)
cp -r product-methodology ~/.claude/skills/
cp -r intern-experience ~/.claude/skills/

# Project-specific only
cp -r product-methodology .claude/skills/
cp -r intern-experience .claude/skills/

Customize

The product-methodology/review-plan-SKILL.md file contains a "Product background" section with placeholders. Fill in your own product context for more targeted reviews:

• Product positioning
• Target market and user personas
• Feature roadmap
• Differences between domestic and overseas versions
• Ops requirements and timeline
• Technical / compliance constraints

License

MIT