夜雨聆风OpenClaw龙虾实战案例 4—— 个人知识库自动整理
场景:收集的文章、笔记、PDF 杂乱无章展示如何用OpenClaw自动收集各平台内容、AI智能分类打标签、统一存储建立索引，并支持与Obsidian/Notion等笔记软件联动，让知识真正"活"起来。
系列教程第 16 篇 | 阅读时长:15-20 分钟
场景背景
知识管理的困境
你是否遇到过这样的问题:
• 收藏了 1000+ 文章,但几乎从不回顾 • 下载了无数 PDF 资料,却找不到想找的那份 • 笔记散落在各个平台:Notion、语雀、印象笔记... • 文件夹结构混乱,整理起来无从下手
问题本质:
• 收集容易,整理难 • 缺乏统一的分类和标签 • 难以检索和关联
自动化目标
用 OpenClaw 实现:
1. 自动收集各平台内容 2. 智能分类和打标签 3. 统一存储到知识库 4. 建立索引便于检索 5. 与笔记软件联动
价值:让知识真正"活"起来,不再沉睡。
一、需求分析:知识库的结构设计
知识库结构
一个良好的知识库应该有清晰的结构:
knowledge_base/├── inbox/ # 收集箱(待整理)│ ├── articles/│ ├── pdfs/│ └── notes/├── projects/ # 项目相关│ ├── project_a/│ └── project_b/├── topics/ # 主题分类│ ├── programming/│ ├── design/│ ├── management/│ └── productivity/├── references/ # 参考资料│ ├── api_docs/│ ├── tutorials/│ └── papers/└── archive/ # 归档 └── 2025/元数据设计
每个知识条目应有以下元数据:
索引设计
创建全文索引,支持快速检索:
{ "index": { "title": "OpenClaw 入门教程", "content_hash": "abc123", "keywords": ["OpenClaw", "AI", "自动化", "教程"], "summary": "本文介绍 OpenClaw 的基本概念和使用方法...", "related": ["article_001", "article_002"], "tags": ["AI", "教程", "自动化"], "created_at": "2026-03-11T10:00:00Z", "path": "/topics/programming/openclaw-tutorial.md" }}二、技能选择:pdf + browser + 自定义整理
本案例需要多个技能配合:
安装技能:
openclaw skill install pdf browser三、自动分类:基于内容的智能归类
分类方法
方法 1:关键词匹配
根据关键词自动分类:
{ "rules": [ { "category": "programming", "keywords": ["代码", "编程", "开发", "Python", "JavaScript"] }, { "category": "design", "keywords": ["设计", "UI", "UX", "视觉", "交互"] }, { "category": "management", "keywords": ["管理", "团队", "协作", "项目"] } ]}方法 2:AI 智能分类
使用 LLM 分析内容,自动判断分类:
{ "classifier": { "type": "llm", "prompt": "分析以下内容,判断它属于哪个主题类别,并给出标签", "categories": ["programming", "design", "management", "productivity", "other"] }}示例:
输入:
标题:OpenClaw 自动化入门内容:OpenClaw 是一个 AI 自动化工具,可以帮助你处理文档...输出:
{ "category": "productivity", "tags": ["AI", "自动化", "效率工具"], "confidence": 0.95}分类流程
新内容 ↓提取文字内容 ↓AI 分析分类 ↓生成标签和摘要 ↓移动到对应目录 ↓更新索引四、标签与索引:便于后续检索
标签系统
标签层级
标签体系├── 领域│ ├── 技术│ │ ├── 前端│ │ ├── 后端│ │ └── AI│ ├── 设计│ └── 管理├── 类型│ ├── 教程│ ├── 案例│ └── 工具└── 状态 ├── 未读 ├── 已读 └── 归档标签规则
{ "tagging_rules": { "auto_tags":true, "min_tags": 2, "max_tags": 5, "allowed_tags": "./tags_list.json" }}索引构建
全文索引
使用 Elasticsearch 或本地索引:
{ "index_config": { "fields": ["title", "content", "tags", "summary"], "analyzer": "chinese", "store_source":true }}检索示例
# 搜索包含"OpenClaw"的内容openclaw knowledge search "OpenClaw"# 按标签筛选openclaw knowledge search --tags "AI,自动化"# 组合查询openclaw knowledge search "教程" --category programming --status unread五、与笔记软件联动
支持的笔记软件
Obsidian 集成示例
Obsidian 基于 Markdown 文件,可以直接集成:
{ "obsidian": { "vault_path": "~/Documents/obsidian/MyVault", "inbox_folder": "Inbox", "auto_frontmatter":true, "frontmatter_template": { "title": "{{title}}", "date": "{{created_at}}", "tags": "{{tags}}", "source": "{{source_url}}", "status": "{{status}}" } }}生成结果:
---title: OpenClaw 自动化入门date: 2026-03-11tags: [AI, 自动化, 教程]source: https://example.com/article/123status: unread---# OpenClaw 自动化入门OpenClaw 是一个 AI 自动化工具...[阅读原文](https://example.com/article/123)Notion 集成示例
通过 Notion API 集成:
{ "notion": { "api_key": "{{env.NOTION_API_KEY}}", "database_id": "xxx-xxx-xxx", "properties_mapping": { "标题": "title", "标签": "multi_select", "来源": "url", "状态": "select", "创建时间": "date" } }}完整工作流配置
创建 knowledge_organizer.json:
{ "name": "知识库自动整理", "description": "自动收集、分类、整理知识内容", "version": "1.0", "trigger": { "type": "manual" }, "inputs": { "source_dir": "./inbox/", "knowledge_base": "./knowledge_base/", "note_app": "obsidian" }, "steps": [ { "id": "step1_scan_new_content", "name": "扫描新内容", "skill": "fs", "command": "scan", "params": { "path": "{{inputs.source_dir}}", "file_types": ["pdf", "md", "html", "txt"] }, "output": "new_files" }, { "id": "step2_extract_content", "name": "提取内容", "skill": "multi", "command": "extract", "params": { "files": "{{new_files}}", "handlers": { "pdf": "pdf.extract-text", "html": "browser.extract-article", "md": "fs.read-text", "txt": "fs.read-text" } }, "output": "raw_contents" }, { "id": "step3_analyze_classify", "name": "分析并分类", "skill": "llm", "command": "classify", "params": { "contents": "{{raw_contents}}", "task": "分析内容,给出分类、标签和摘要", "output_format": { "title": "string", "category": "string", "tags": "array", "summary": "string(200字以内)", "importance": "high|medium|low" } }, "output": "analyzed_data" }, { "id": "step4_determine_path", "name": "确定存储路径", "skill": "custom", "command": "determine-path", "params": { "category": "{{analyzed_data.category}}", "importance": "{{analyzed_data.importance}}", "date": "{{date}}", "base_path": "{{inputs.knowledge_base}}" }, "output": "target_paths" }, { "id": "step5_organize_files", "name": "整理文件", "skill": "fs", "command": "organize", "params": { "files": "{{new_files}}", "target_paths": "{{target_paths}}", "add_frontmatter":true, "frontmatter": { "title": "{{analyzed_data.title}}", "tags": "{{analyzed_data.tags}}", "summary": "{{analyzed_data.summary}}", "date": "{{date}}", "status": "unread" } }, "output": "organized_files" }, { "id": "step6_update_index", "name": "更新索引", "skill": "search", "command": "index", "params": { "files": "{{organized_files}}", "index_path": "{{inputs.knowledge_base}}/index.json", "fields": ["title", "tags", "summary", "content"] } }, { "id": "step7_sync_to_notes", "name": "同步到笔记软件", "skill": "{{inputs.note_app}}", "command": "sync", "params": { "files": "{{organized_files}}", "sync_mode": "add" }, "condition": "{{inputs.note_app != 'none'}}" }, { "id": "step8_generate_report", "name": "生成整理报告", "skill": "xlsx", "command": "create", "params": { "data": "{{analyzed_data}}", "file": "./reports/knowledge_report_{{date}}.xlsx", "columns": ["标题", "分类", "标签", "重要程度", "状态", "路径"] } } ], "outputs": { "report": "./reports/knowledge_report_{{date}}.xlsx", "indexed": "{{organized_files.count}}" }}执行效果
执行命令
openclaw workflow run knowledge_organizer.json执行日志
执行工作流:知识库自动整理─────────────────────────────[1/8] 扫描新内容... ✓ 发现 15 个新文件 - PDF: 8 个 - 文章: 5 个 - 笔记: 2 个[2/8] 提取内容... ✓ 已提取 15 个文件的内容[3/8] 分析并分类... 使用 AI 分析内容... ✓ 分析完成 分类结果: ├─ programming: 5 篇 ├─ design: 3 篇 ├─ productivity: 4 篇 └─ management: 3 篇[4/8] 确定存储路径... ✓ 路径已确定[5/8] 整理文件... ✓ 已整理并移动文件 ✓ 已添加元数据[6/8] 更新索引... ✓ 索引已更新 当前索引条目: 1,234 条[7/8] 同步到笔记软件... 同步至 Obsidian... ✓ 已同步 15 条到 Obsidian[8/8] 生成整理报告... ✓ 已生成报告─────────────────────────────✓ 工作流执行成功!处理结果: 新增内容: 15 条 索引总数: 1,234 条输出文件: - knowledge_report_2026-03-11.xlsx已同步至 Obsidian整理结果示例
原始文件:
inbox/├── article_001.pdf├── article_002.html└── note_003.txt整理后:
knowledge_base/├── topics/│ ├── programming/│ │ ├── 2026-03-11_python-best-practices.md│ │ └── 2026-03-11_web-performance.md│ └── productivity/│ └── 2026-03-11_time-management-tips.md└── references/ └── papers/ └── 2026-03-11_ai-research.pdfMarkdown 文件内容:
---title: Python 最佳实践指南tags: [Python, 编程, 最佳实践]summary: 本文总结了 Python 开发中的最佳实践,包括代码风格、性能优化等date: 2026-03-11status: unreadimportance: highsource: https://example.com/python-best-practices---# Python 最佳实践指南[原文内容...]## 关键要点1. 遵循 PEP 8 代码风格2. 使用类型注解提高代码可读性3. 编写单元测试保证质量小结
本案例展示了如何用 OpenClaw 实现知识库自动整理:
1. ✅ 自动扫描和收集内容 2. ✅ AI 智能分类和打标签 3. ✅ 自动整理文件结构 4. ✅ 建立全文索引 5. ✅ 与笔记软件联动
通过自动化整理,知识不再沉睡,真正成为你的资产。
下篇预告
实战篇到此结束,下一篇我们进入高级篇,学习如何编写自己的技能。
下一篇:如何编写自己的技能?