home / skills / alchaincyf / huashu-skills / huashu-prompt-save

huashu-prompt-save skill

safe

This skill automatically classifies prompts into five categories, saves them with structured Markdown, and updates an index for quick retrieval.

npx playbooks add skill alchaincyf/huashu-skills --skill huashu-prompt-save

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

2.4 KB

---
name: huashu-prompt-save
description: 自动识别Prompt类型并分类保存（技术/内容/教学/产品/通用）。当用户提到"保存prompt"、"记录prompt"、"整理prompt"时使用。
---

# Prompt分类保存

自动识别Prompt类型、保存到对应分类、更新索引的专业能力。

## 何时使用

- 用户说"保存这个prompt"、"记录一下"、"存起来"
- 用户粘贴prompt并说"帮我整理"

## 核心功能

1. **自动识别类型**：5大分类（技术开发/内容创作/教学设计/产品咨询/通用工具）
2. **提取信息**：用途、来源、适用场景、关键特点
3. **自动保存**：`对应分类/prompt-[序号]-[标题].md`
4. **更新索引**：在 `INDEX.md` 中添加条目

## 5大分类规则

| 分类 | 关键词 | 示例 |
|-----|--------|------|
| 技术开发类 | 编程、调试、代码、优化 | "Python代码优化Prompt" |
| 内容创作类 | 写作、文案、脚本、创作 | "公众号文章生成器" |
| 教学设计类 | 课程、教学、讲解、训练 | "知识点拆解Prompt" |
| 产品咨询类 | 需求、产品、用户、增长 | "用户需求分析Prompt" |
| 通用工具类 | 通用、效率、助手 | "通用翻译工具" |

## 保存格式

```markdown
# [一句话标题]

**来源**: 自创 / 网络 / 修改自xxx
**标签**: #标签1 #标签2 #标签3
**适用场景**: xxx

## Prompt 内容
[完整 prompt，保持原格式]

## 使用示例
[如果有示例或使用心得]

## 效果/备注
[使用效果、注意事项]
```

## 执行流程

1. **识别类型**：根据关键词和内容判断分类
2. **提取信息**：用途、来源、标签、适用场景
3. **生成文件名**：`prompt-[自动递增序号]-[简短标题].md`
4. **保存文件**：到 `/Prompt梳理/[分类]/` 目录
5. **更新索引**：在 `/Prompt梳理/INDEX.md` 中添加条目

## 注意事项

- 如果分类不明确，询问用户意图
- 标题要实用，一眼看懂用途（不超过15字）
- 标签用于快速检索，3-5个为宜
- 保持原prompt的格式和结构
- 网络来源的尽量记录出处

## 参考资源

- `/Prompt梳理/CLAUDE.md` - Prompt管理规则

**最后更新**: 2025-11-07

---

> **花叔出品** | AI Native Coder · 独立开发者
> 公众号「花叔」| 30万+粉丝 | AI工具与效率提升
> 代表作：小猫补光灯（AppStore付费榜Top1）·《一本书玩转DeepSeek》

Overview

This skill automatically recognizes prompt type and saves it into one of five categories (Technical, Content, Teaching, Product, General). It extracts key metadata, generates a clear filename and markdown file, and updates the index for easy retrieval. The goal is quick, consistent prompt organization so users can store and reuse prompts reliably.

How this skill works

When the user asks to save or organize a prompt, the skill scans the text for keywords and intent to assign one of five categories. It extracts metadata (purpose, source, applicable scenarios, tags), preserves the original prompt format, and generates a concise title and filename. The skill writes a structured markdown entry into the matching category folder and appends an entry to the central INDEX.md.

When to use it

User says: "save this prompt", "record prompt", or "整理prompt" (organize prompt).
User pastes a prompt and asks: "help me store/organize this."
You want consistent naming, tagging, and searchable storage for reused prompts.
You are creating a shareable prompt library for a team or project.

Best practices

If intent is ambiguous, ask a clarifying question before saving.
Keep titles practical and under 15 characters for quick scanning.
Assign 3–5 descriptive tags to improve searchability.
Preserve the original prompt formatting exactly in the saved file.
Record network sources when the prompt is adapted from external content.

Example use cases

Saving a Python debugging prompt into the Technical category with tags like #python #debugging.
Organizing a blog-post generator prompt into Content with usage notes and example outputs.
Storing a lesson-plan prompt into Teaching with target audience and suggested activities.
Recording a product-research prompt into Product with user-research context and recommended metrics.
Collecting small utility prompts (translation, concise summaries) into General for team utilities.

FAQ

What if the prompt fits multiple categories?

Ask the user which category they prefer; if unsure, recommend the category matching the primary intent and allow adding cross-reference tags.

How are filenames generated?

Filenames use an auto-incremented number plus a short practical title: prompt-[number]-[short-title].md to keep entries ordered and readable.