From yuque-personal
Refines scattered Yuque notes into organized, deduplicated, enriched documents via yuque-mcp tools. Searches repos/docs, analyzes themes/duplicates, restructures for personal knowledge bases.
npx claudepluginhub yuque/yuque-pluginThis skill uses the workspace's default tool permissions.
Help the user transform scattered, rough notes into well-structured, deduplicated, and enriched documents. Turn chaos into clarity.
Refines rough notes into structured documents with headings, improved phrasing, and formatting via Yuque tools. Handles links or pasted text; preserves original meaning for personal use.
Builds structured Obsidian notes from raw content with entity extraction, wikilinks, Zettelkasten atomic formatting, meeting notes, and knowledge graph integration.
Builds and maintains persistent Obsidian wiki vaults using AI for source ingestion, knowledge querying, note linting, and autonomous research.
Share bugs, ideas, or general feedback.
Help the user transform scattered, rough notes into well-structured, deduplicated, and enriched documents. Turn chaos into clarity.
All tools are from the yuque-mcp server:
yuque_search — Find the target notes to refineyuque_get_doc — Read the full content of notesyuque_list_repos — List personal reposyuque_list_docs — List documents in a repoyuque_create_doc — Create the refined documentyuque_update_doc — Update the original document in placeThe user may provide:
If the user says "整理我最近的笔记":
Tool: yuque_search
Parameters:
query: "每日捕获"
type: "doc"
Or for topic-specific notes:
Tool: yuque_search
Parameters:
query: "<topic keyword>"
type: "doc"
For each identified document:
Tool: yuque_get_doc
Parameters:
repo_id: "<namespace>"
doc_id: "<slug>"
Read up to 5-10 source documents. If there are more, prioritize by recency and relevance.
Before refining, analyze the raw content:
Present the analysis to the user:
## 📋 笔记分析
**来源**:X 篇文档,共约 XXXX 字
### 主题分布
- **[主题 1]**:X 条相关记录
- **[主题 2]**:X 条相关记录
- **[主题 3]**:X 条相关记录
### 发现
- 🔄 发现 X 处重复内容
- 🕳️ 发现 X 处信息缺口
- 📊 建议整理为 X 篇主题文档
要按这个方向整理吗?
Based on user confirmation, perform the refinement:
Restructure one document:
Merge multiple notes into themed documents:
Refined document template:
# [主题标题]
> 整理自 X 篇笔记,时间跨度:YYYY-MM-DD 至 YYYY-MM-DD
> 最后整理:YYYY-MM-DD
---
## 概述
[对这个主题的整体概述,2-3 句话]
---
## [分类 1]
### [子主题 1.1]
[整理后的内容]
### [子主题 1.2]
[整理后的内容]
---
## [分类 2]
### [子主题 2.1]
[整理后的内容]
---
## 待补充
- [ ] [识别出的信息缺口 1]
- [ ] [识别出的信息缺口 2]
---
## 原始来源
- [原文档 1 标题](链接) — YYYY-MM-DD
- [原文档 2 标题](链接) — YYYY-MM-DD
---
> 本文档由 AI 助手整理,原始笔记已保留。
Ask the user's preference:
For new document:
Tool: yuque_create_doc
Parameters:
repo_id: "<namespace>"
title: "[主题标题]"
body: "<refined content>"
format: "markdown"
For in-place update:
Tool: yuque_update_doc
Parameters:
repo_id: "<namespace>"
doc_id: "<slug>"
body: "<refined content>"
✅ 笔记整理完成!
📄 **[[主题标题]](文档链接)**
📚 已保存到:「知识库名称」
### 整理成果
- 📥 输入:X 篇原始笔记,约 XXXX 字
- 📤 输出:1 篇结构化文档,约 XXXX 字
- 🔄 去除重复:X 处
- 🕳️ 标记待补充:X 处
💡 原始笔记已保留,可随时对照。
| Situation | Action |
|---|---|
| Source notes are too few (<3 items) | Refine what's available; suggest capturing more first |
| Source notes are too many (>10 docs) | Focus on most recent or ask user to narrow the scope |
| Content is already well-structured | Tell user honestly; suggest minor improvements only |
yuque_update_doc fails | Fall back to creating a new document |
| Mixed languages in notes | Maintain the dominant language; don't force translation |