Obsidian MCP Server
Enables AI assistants to read, search, create, and organize notes in an Obsidian vault, with features like summarization and daily most valuable notes identification.
README
Obsidian MCP Server
A Model Context Protocol (MCP) server that enables AI assistants to interact with your Obsidian vault.
Features
1. Note Reading & Summarization
- Read any note in your vault
- Generate quick summaries of notes
- Extract metadata (tags, word count, dates)
2. Note Organization
- Move notes between folders
- Add tags to notes automatically
- Search notes by title, content, or tags
3. Daily "Most Valuable Notes"
- Automatically identifies your most important notes based on:
- Recent modifications (last 7 days)
- Note length and depth
- Tag organization
- Perfect for daily reviews and prioritization
Installation
1. Create Project Directory
mkdir obsidian-mcp-server
cd obsidian-mcp-server
2. Initialize Project
Copy the following files into your project:
src/index.ts(the main server code)package.jsontsconfig.json
3. Install Dependencies
npm install
4. Build the Server
npm run build
Configuration
Environment Variables
Set the path to your Obsidian vault:
macOS/Linux:
export OBSIDIAN_VAULT_PATH="/path/to/your/obsidian/vault"
Windows (PowerShell):
$env:OBSIDIAN_VAULT_PATH="C:\path\to\your\obsidian\vault"
Windows (CMD):
set OBSIDIAN_VAULT_PATH=C:\path\to\your\obsidian\vault
Claude Desktop Configuration
Add this to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Roaming\Claude\claude_desktop_config.json
{
"mcpServers": {
"obsidian": {
"command": "node",
"args": ["/absolute/path/to/obsidian-mcp-server/build/index.js"],
"env": {
"OBSIDIAN_VAULT_PATH": "/path/to/your/obsidian/vault"
}
}
}
}
Available Tools
1. read_note
Read the content of a specific note.
{
"path": "folder/note-name.md"
}
2. search_notes
Search for notes by title, content, or tags.
{
"query": "machine learning",
"searchInContent": true
}
3. create_note
Create a new note in your vault.
{
"path": "folder/new-note.md",
"content": "# New Note\n\nContent here..."
}
4. summarize_note
Generate a summary of a note.
{
"path": "folder/long-note.md"
}
5. organize_note
Move a note to a different folder.
{
"path": "old-folder/note.md",
"targetFolder": "new-folder"
}
6. add_tags
Add tags to a note's frontmatter.
{
"path": "folder/note.md",
"tags": ["important", "project-x"]
}
7. get_most_valuable_notes
Get your top 10 most valuable notes for today.
{}
Usage Examples
Once configured, you can ask Claude in the desktop app:
- "Show me my most valuable notes for today"
- "Search my vault for notes about Python"
- "Summarize my note called 'Project Ideas'"
- "Move my note 'Draft.md' to the 'Archive' folder"
- "Add tags 'urgent' and 'work' to my note 'Meeting Notes.md'"
- "Create a new note in 'Daily Notes' called 'Today.md' with today's date"
Development
Watch Mode
For development, run TypeScript in watch mode:
npm run dev
Testing
Test the server directly:
export OBSIDIAN_VAULT_PATH="/path/to/vault"
npm start
Troubleshooting
Common Issues
"OBSIDIAN_VAULT_PATH environment variable not set"
- Make sure you've set the environment variable correctly
- Use absolute paths, not relative paths
- Check for typos in the path
"Permission denied" errors
- Ensure the build/index.js file is executable:
chmod +x build/index.js - Check that Node.js has read/write permissions to your vault
Claude Desktop doesn't see the server
- Restart Claude Desktop after editing the config file
- Check the config file syntax (valid JSON)
- Use absolute paths in the config file
- Check Claude Desktop logs for errors
Future Enhancements
Potential features to add:
- Link analysis and graph visualization
- Automated note templates
- Batch operations
- Daily note automation
- Integration with calendar and tasks
- AI-powered note suggestions
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。