Logseq MCP Server
Enables AI assistants like Claude to directly read, write, search, and navigate your local Logseq knowledge graph, including managing journals, pages, backlinks, and page relationships without manual copy-pasting.
README
Logseq MCP Server
Let AI read and write your Logseq graph directly via MCP
Talk to Claude and say "add this to today's journal", "find what I did last week", "show me all pages linked to this one" - and it just works.
Why This?
Problem: Logseq is a great PKM tool, but integrating with AI assistants requires constant copy-pasting.
Solution: With this MCP server:
- Claude directly writes to your journal (no copy-paste)
- Search and summarize past entries (maintain context)
- Navigate connections between pages (backlinks, graph)
- Auto-generate daily journals with templates
You: "Summarize today's meeting notes and add them to my journal"
Claude: [writes directly to Logseq via logseq-mcp]
"Done! Added to today's journal. Anything else?"
Is This For You?
Good fit if you...
- Use Logseq as your primary PKM
- Use Claude Code or Claude Desktop regularly
- Want to delegate note management to AI
- Use local file-based Logseq (not Logseq Sync)
Not for you if...
- Using Logseq Sync (requires local file access)
- Obsidian user (different MCP server needed)
- Have sensitive info in notes and uncomfortable with AI access
- Use org-mode instead of Markdown (not yet supported)
Features
| Feature | Description |
|---|---|
| Page CRUD | Create, read, update, delete pages + property support |
| Search | Full-text search + tag/folder filtering |
| Graph Navigation | Links, backlinks, page relationship traversal |
| Journal | Access today's/specific date journals + templates |
| Content Logging | Log articles, books, movies, exhibitions to journal |
| Resources | Expose graph pages as MCP resources |
Quick Start
1. Install
git clone https://github.com/dearcloud09/logseq-mcp.git
cd logseq-mcp
npm install
npm run build
2. Configure
Claude Code (~/.claude/settings.json):
{
"mcpServers": {
"logseq": {
"command": "node",
"args": ["/path/to/logseq-mcp/dist/index.js"],
"env": {
"LOGSEQ_GRAPH_PATH": "/path/to/your/logseq/graph"
}
}
}
}
Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"logseq": {
"command": "node",
"args": ["/path/to/logseq-mcp/dist/index.js"],
"env": {
"LOGSEQ_GRAPH_PATH": "/path/to/your/logseq/graph"
}
}
}
}
3. Verify
Ask Claude: "Show me my Logseq page list"
Available Tools
| Tool | Description |
|---|---|
list_pages |
List all pages with metadata (tags, links, backlinks) |
read_page |
Read page content and metadata |
create_page |
Create new page (with property support) |
update_page |
Update page content |
delete_page |
Delete a page |
append_to_page |
Append content to existing page |
search_pages |
Search by content/title + tag/folder filters |
get_backlinks |
Get pages that reference a specific page |
get_graph |
Get page connection graph data |
get_journal |
Get today's or specific date's journal |
create_journal |
Create journal with optional template |
add_article |
Add article to journal (title, summary, tags, URL, highlights) |
add_book |
Add book to journal (title, author, tags, memo) |
add_movie |
Add movie to journal (title, director, memo) |
add_exhibition |
Add exhibition to journal (title, venue, artist, memo) |
Usage Examples
"Show me today's journal"
"Add this content to 'Project A' page: ..."
"Find all pages with #meeting tag"
"What pages are connected to my Goals page?"
"Search for TODO items in last week's journals"
"Create a new page called 'Reading List'"
"Summarize our conversation and save it as an article in my journal"
Logseq Graph Structure
your-graph/
journals/ # Daily journals (2024_01_15.md format)
pages/ # Regular pages
logseq/ # Logseq settings
whiteboards/ # Whiteboards
Security
- Path traversal protection (graph-only access)
- Symlink/hardlink attack prevention
- Input validation and sanitization
- DoS protection (content size limits)
- Error message sanitization
Troubleshooting
"LOGSEQ_GRAPH_PATH environment variable is required"
Set LOGSEQ_GRAPH_PATH in your configuration file.
MCP server not recognized by Claude
- Restart Claude Code/Desktop
- Verify path is absolute (
/Users/...format) - Ensure
npm run buildwas executed
Pages not showing up
- Check if
.mdfiles exist injournals/orpages/ - Verify you're using local graph (not Logseq Sync)
org-mode files not reading
Currently Markdown only. org-mode support planned for future.
Korean-Specific Features
This project includes features optimized for Korean users:
Daily Automation (Optional)
Auto-generate daily journal with weather (Korea only - uses Naver Weather) and diary template:
LOGSEQ_GRAPH_PATH=/path/to/logseq WEATHER_LOCATION=서울 node add-today-dairy.js
See Korean README for detailed setup with launchd automation.
Cultural Content Structure
add_book, add_movie, add_exhibition tools use Korean wikilink structure ([[문화]]). Customize the templates in src/index.ts for your language.
Development
# Development mode (watch)
npm run dev
# TypeScript build
npm run build
# Production run
npm start
Project Structure
src/
index.ts # MCP server entry point, tool handlers
types.ts # TypeScript type definitions
graph.ts # Graph filesystem operations
Contributing
Issues and PRs welcome!
- Fork this repo
- Create feature branch (
git checkout -b feature/amazing) - Commit changes (
git commit -m 'Add amazing feature') - Push to branch (
git push origin feature/amazing) - Open a Pull Request
Ideas for contribution
- [ ] org-mode support
- [ ] Logseq property search
- [ ] Whiteboard support
- [ ] Better graph visualization data
- [ ] i18n for templates
License
Polyform Noncommercial 1.0.0 - Free for personal and noncommercial use.
Related
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。