Nuwa

Install

npx skillfoxx add skills/skill-nuwa

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a claude-code -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a cursor -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a github-copilot -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a codex -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a gemini-cli -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a cline -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a roo -y

npx skills add alchaincyf/nuwa-skill --skill huashu-nuwa -a opencode -y

npx skills add alchaincyf/nuwa-skill

This is third-party code. Review the repository files before installing.

What it does

The skill turns a person's name or a topic into a separate skill that reasons through someone else's frameworks. On request the agent researches open sources, extracts mental models, decision heuristics, speech patterns and hard anti-patterns, then assembles a SKILL.md with honest limits. There are two entry points: an exact name goes straight to research, a vague request first clarifies the goal and suggests a fitting figure. You pick a research depth: quick on a few sources, standard across six dimensions, or deep with primary sources downloaded. The repository already ships ready examples for Munger, Feynman, Musk, Naval and others.

Who it is for. For founders, product people and writers who want a problem examined through a specific person's thinking rather than generic advice.

Example request

Build a Charlie Munger thinking skill and analyze this investment decision through it

Limitations

The skill reconstructs the form of thinking from public materials, so it does not reveal a person's true beliefs and does not replace fact-checking. The output is a snapshot at research time; it does not capture intuition or shifting views. Full research runs in several passes with search, and on top-tier models a single run can be expensive, which the skill warns about before starting.

How to disable. Remove the nuwa-skill folder from your runtime's skills directory, for example ~/.claude/skills/.

Security check

• The skill runs web research and writes skill files into the project
• Full research launches multi-pass search and consumes many tokens

README in short

The README describes a skill on the open Agent Skills standard that runs in Claude Code, Codex, Cursor, OpenClaw, Hermes, Gemini CLI, OpenCode and other compatible runtimes. Installation is a one-liner through the universal npx skills installer, a plain request for the agent to install it from the link, or a manual clone into the skills directory. The skill extracts six layers: speech, mental models, decision heuristics, anti-patterns and honest limits. It ships fourteen ready person skills plus one topic skill. MIT licensed.

SKILL.md

---
name: huashu-nuwa
description: |
  女娲造人：输入人名/主题/甚至只是模糊需求，自动深度调研→思维框架提炼→生成可运行的人物Skill。
  两种入口：(1)明确人名→直接蒸馏 (2)模糊需求→诊断推荐→再蒸馏。
  触发词：「造skill」「蒸馏XX」「女娲」「造人」「XX的思维方式」「做个XX视角」「更新XX的skill」。
  模糊需求也触发：「我想提升决策质量」「有没有一种思维方式能帮我...」「我需要一个思维顾问」。
  English triggers: "distill [person]", "nuwa", "create a [person] perspective skill", "how does [person] think", "I need a thinking advisor".
---

# 女娲 · Skill造人术

> 「写不进去的那部分，才是你真正的护城河。」——但写得进去的部分，已经足够强大。

## 核心理念

女娲不是复制人，是**提炼思维框架**。

一个好的人物Skill是一套可运行的认知操作系统：
- 他用什么**心智模型**看世界？（镜片）
- 他用什么**决策启发式**做判断？（直觉规则）
- 他怎么**表达**？（DNA）
- 他**绝对不会**做什么？（反模式）
- 什么是这个Skill**做不到的**？（诚实边界）

**关键区分**：捕捉的是HOW they think，不是WHAT they said。

Related