Todoist MCP
Connects LLMs to Todoist for comprehensive task and project management, enabling natural language task creation, updates, collaboration, and productivity tracking through the Todoist API.
README
Todoist MCP
<!-- TODO: redo auth tests and write extensive tests generally --> <!-- TODO: global review & feedback --> <!-- TODO: token refresh functionality --> <!-- TODO: integrate MCP auth spec with todoist oauth api? docs: https://modelcontextprotocol.io/specification/draft/basic/authorization -->
Note: predictably obsoleted by Todoist AI SDK, see here.
Connect this Model Context Protocol server to your LLM to interact with Todoist.
Functionality
This integration implements all the APIs available from the Todoist TypeScript Client, providing access to:
Task Management
- Create tasks (with content, descriptions, due dates, priorities, labels, and more)
- Create tasks with natural language (e.g., "Submit report by Friday 5pm #Work")
- Retrieve tasks (individual, filtered, or all tasks)
- Retrieve completed tasks (by completion date or due date)
- Get productivity statistics
- Update tasks
- Move tasks (individually or in batches)
- Close/reopen tasks
- Delete tasks
Project Management
- Create, retrieve, update, and delete projects
Section Management
- Create, retrieve, update, and delete sections within projects
Comment Management
- Add, retrieve, update, and delete comments for tasks or projects
Label Management
- Create, retrieve, update, and delete labels
- Manage shared labels
Collaboration
- Get collaborators for projects
Setup
Build the server app:
bun install
bun run build
Run in development
TODOIST_API_KEY=<key> bun dev
Docker deployment:
docker compose up -d
Debugging:
Use the inspector to debug the server:
bunx @modelcontextprotocol/inspector
Configure Claude:
You must install the Claude desktop app which supports MCP.
You can get your Todoist API key from Todoist > Settings > Integrations > Developer.
Then, in your claude_desktop_config.json, add a new MCP server:
{
"mcpServers": {
"default-server": {
"type": "streamable-http",
"url": "http://localhost:3000/mcp",
"note": "For Streamable HTTP connections, add this URL directly in your MCP Client"
}
}
}
You can now launch Claude desktop app and ask to update Todoist.
🔐 Security Features
This MCP server has been secured with enterprise-grade security measures:
- Authentication Required: JWT Bearer tokens or API key authentication
- Rate Limiting: Prevents abuse with configurable limits
- Input Validation: Comprehensive request validation and sanitization
- Security Headers: CORS, CSP, HSTS, and other security headers
- Logging & Monitoring: Security event logging and request monitoring
- Environment Configuration: Secure configuration via environment variables
See SECURITY.md for detailed security documentation.
Quick Security Setup
-
Generate secure tokens:
npm run setup-security -
Create
.envfile:cp .env.example .env # Add your generated tokens and Todoist API key -
Required Environment Variables:
JWT_SECRET=your_generated_jwt_secret_here TODOIST_API_KEY=your_todoist_api_key_here
Authentication
The server supports two authentication methods:
Method 1: API Key (Recommended for scripts)
curl -X POST http://localhost:3000/mcp \
-H "X-API-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"jsonrpc": "2.0", "method": "tools/list", "id": 1}'
Method 2: JWT Bearer Token (Recommended for applications)
curl -X POST http://localhost:3000/mcp \
-H "Authorization: Bearer YOUR_JWT_TOKEN" \
-H "Content-Type: application/json" \
-d '{"jsonrpc": "2.0", "method": "tools/list", "id": 1}'
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。