warm handover

name: warm-handover
description: "Transform a colleague departure into a structured, warm, and living knowledge transfer. The human alternative to colleague-skill. Use when: (1) someone is leaving the team and needs to hand over their work, (2) a new person is taking over an existing project, (3) you want to capture tribal knowledge before it walks out the door, (4) a team member is transitioning roles internally. NOT for: scraping chat logs to create an AI clone of a person."

Warm Handover

"冰冷的离别化为温暖的 Skill" — 但真正的温暖不是模仿死人，而是让活人好好交接。

colleague-skill 的做法：爬聊天记录 → 生成 AI 替身 → 假装他没走。 Warm Handover 的做法：引导式访谈 → 结构化知识 → 让接手者真正能干活。

How It Works

Phase 1: Guided Interview

Run a structured handover interview with role-based templates:

# CLI mode
python3 scripts/handover_interview.py

# Web mode
cd web && npm run dev
# Open http://localhost:3000

Phase 2: Structured Output

Generates four files:

• PROJECT.md — 项目全景、架构、关键决策
• PEOPLE.md — 人脉地图、协作网络
• LESSONS.md — 经验教训、踩坑记录
• TODO.md — 待办清单、已知问题

Phase 3: Living Document

• 接手者每遇到新问题，更新对应文件
• 每月回顾 LESSONS.md，把个人经验变成团队知识
• 3 个月后评估交接质量

Why This Beats Colleague-Skill

Philosophy

colleague-skill 的核心假设是：一个人的价值 = 他的聊天记录。 这是对人最大的不尊重。

Warm Handover 不制造 AI 替身。它帮助一个人好好告别，帮助另一个人好好开始。

References

• references/interview-templates.md — 不同角色的定制化访谈模板
• references/handover-checklist.md — 完整的交接清单和时间线管理