joplin-mcp-server
Joplin MCP Server
README
Joplin MCP Server
This is a Node.js implementation of an MCP (Model Context Protocol) server for Joplin.
Installation
npm install
Configuration
Create a .env file with the following variables:
JOPLIN_PORT=41184
JOPLIN_TOKEN=your_joplin_token
You can find your Joplin token in the Joplin desktop app under: Tools > Options > Web Clipper
Usage
Local Development
Start the server:
npm start
You can also specify a custom environment file:
npm start -- --env-file .env.custom
Using npx (Recommended)
After publishing to npm, you can use npx to run the server without installation:
# Using command line arguments
npx joplin-mcp-server --port 41184 --token your_joplin_token
# Using environment file
npx joplin-mcp-server --env-file /path/to/your/.env
# Mixed approach (args override env file)
npx joplin-mcp-server --env-file .env --port 41185
Command Line Options
OPTIONS:
--env-file <file> Load environment variables from file
--port <port> Joplin port (default: 41184)
--token <token> Joplin API token
--help, -h Show help message
MCP Client Configuration
Usage in Augment Code:
name: joplin
command: npx joplin-mcp-server --port 41184 --token your_token
Usage in mcp.json (Cursor and other tools):
{
"joplin": {
"command": "npx",
"args": ["joplin-mcp-server", "--port", "41184", "--token", "your_joplin_token"]
}
}
Or using environment file:
{
"joplin": {
"command": "npx",
"args": ["joplin-mcp-server", "--env-file", "/path/to/your/.env"]
}
}
Legacy Usage (if installed locally)
Usage in Augment Code:
name: joplin
command: node /path/to/your/mcp-joplin/index.js --env-file /path/to/your/mcp-joplin/.env
Usage in mcp.json (cursor other tools)
"joplin":{
"command":"node",
"args":[
"/path/to/your/mcp-joplin/index.js",
"--env-file",
"/path/to/your/mcp-joplin/.env"
]
}
Logging
The server logs all incoming commands and outgoing responses. Logs are stored in two places:
- Console output: Basic information is displayed in the console
- Log files: Detailed logs are saved in the
logsdirectory with timestamps
You can adjust the log level by setting the LOG_LEVEL environment variable:
LOG_LEVEL=debug npm start
Available log levels (from most to least verbose):
debug: All messages including detailed command and response datainfo: Standard operational messages (default)warn: Warnings and errors onlyerror: Only error messages
Available Tools
list_notebooks
Retrieves the complete notebook hierarchy from Joplin.
# Example output:
Notebook 1 (id: "abc123")
Subnotebook 1.1 (id: "def456")
Subnotebook 1.2 (id: "ghi789")
Notebook 2 (id: "jkl012")
search_notes
Searches for notes in Joplin and returns matching notebooks.
Parameters:
query: The search query string
# Example usage:
search_notes query="project meeting"
# Example output:
Found 2 notes matching query: "project meeting"
NOTE: To read a notebook, use the notebook ID (not the note title)
- Note: "Weekly Project Meeting" (note_id: "abc123")
Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Updated: 3/15/2025, 10:30:45 AM
Snippet: Notes from our weekly project meeting. Topics discussed: timeline, resources, next steps...
To read this notebook: read_notebook notebook_id="58a0a29f68bc4141b49c99f5d367638a"
- Note: "Project Kickoff Meeting" (note_id: "def456")
Notebook: "Projects" (notebook_id: "72b1c45d89ef3212a67b98f4e5d23a1b")
Updated: 3/10/2025, 2:15:30 PM
Snippet: Initial project meeting with stakeholders. Key decisions: project scope, team members...
To read this notebook: read_notebook notebook_id="72b1c45d89ef3212a67b98f4e5d23a1b"
Important: Note the difference between note titles and IDs. When using the
read_notebookcommand, you must use the notebook ID (a long alphanumeric string), not the notebook title.
read_notebook
Reads the contents of a specific notebook.
Parameters:
notebook_id: The ID of the notebook to read
# Example usage:
read_notebook notebook_id="58a0a29f68bc4141b49c99f5d367638a"
# Example output:
# Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Contains 3 notes:
NOTE: This is showing the contents of notebook "Work", not a specific note.
- Note: "Weekly Project Meeting" (note_id: "def456")
Updated: 3/15/2025, 10:30:45 AM
- ✅ Note: "Call client" (note_id: "ghi789")
Updated: 3/14/2025, 3:45:12 PM
- ☐ Note: "Prepare presentation" (note_id: "jkl012")
Updated: 3/13/2025, 9:20:33 AM
Common Error: If you try to use a note title (like "todo") instead of a notebook ID, you'll get an error. Always use the notebook ID (the long alphanumeric string) shown in the search results or notebook list.
read_note
Reads the full content of a specific note.
Parameters:
note_id: The ID of the note to read
# Example usage:
read_note note_id="def456"
# Example output:
# Note: "Weekly Project Meeting"
Note ID: def456
Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Created: 3/15/2025, 10:00:12 AM
Updated: 3/15/2025, 10:30:45 AM
---
# Weekly Project Meeting
## Agenda
1. Project status update
2. Timeline review
3. Resource allocation
4. Next steps
## Notes
- Project is on track for Q2 delivery
- Need to allocate additional resources to the UI team
- Next meeting scheduled for next Friday
---
Related commands:
- To view the notebook containing this note: read_notebook notebook_id="58a0a29f68bc4141b49c99f5d367638a"
- To search for more notes: search_notes query="your search term"
Note: The
read_notecommand shows the full content of a specific note, while theread_notebookcommand shows a list of notes in a notebook. Usesearch_notesto find notes and get their IDs.
read_multinote
Reads the full content of multiple notes at once.
Parameters:
note_ids: An array of note IDs to read
# Example usage:
read_multinote note_ids=["def456", "ghi789", "jkl012"]
# Example output:
# Reading 3 notes
## Note 1 of 3 (ID: def456)
### Note: "Weekly Project Meeting"
Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Created: 3/15/2025, 10:00:12 AM
Updated: 3/15/2025, 10:30:45 AM
---
# Weekly Project Meeting
## Agenda
1. Project status update
2. Timeline review
---
## Note 2 of 3 (ID: ghi789)
### Note: "Call client"
Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Status: Completed
Created: 3/14/2025, 3:00:00 PM
Updated: 3/14/2025, 3:45:12 PM
---
Discussed project timeline and next steps.
Client is happy with progress.
---
## Note 3 of 3 (ID: jkl012)
### Note: "Prepare presentation"
Notebook: "Work" (notebook_id: "58a0a29f68bc4141b49c99f5d367638a")
Status: Not completed
Due: 3/20/2025, 9:00:00 AM
Created: 3/13/2025, 9:00:00 AM
Updated: 3/13/2025, 9:20:33 AM
---
# Presentation Outline
- Introduction
- Project overview
- Timeline
- Budget
- Next steps
---
# Summary
Total notes requested: 3
Successfully retrieved: 3
Tip: When you search for notes or view a notebook, you'll see a suggestion for using
read_multinotewith the exact IDs of the notes found. This makes it easy to read multiple related notes at once.
Development
Local Development Setup
To test the npx command locally during development:
# In the project root directory
cd /path/to/mcp-joplin
npm install
npm link
After linking, you can test your local changes immediately:
# Test the CLI
npx joplin-mcp-server --help
npx joplin-mcp-server --port 41184 --token your_token
# Make code changes, then test again (no rebuild needed)
npx joplin-mcp-server --help
Making Changes
- Edit any
.jsfiles in the project - Run tests:
npm test - Test the CLI:
npx joplin-mcp-server --help
No build step is required - changes are immediately available through the npm link.
Running Tests
Create a .env.test.local file with your test configuration, then run:
npm test
Publishing to npm
To make this package available via npx:
- Update the version in
package.json - Run
npm publish
Users can then run it with npx joplin-mcp-server
Unlinking (if needed)
To remove the local link:
npm unlink -g joplin-mcp-server
License
MIT
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。