Todoist MCP Server
MCP server that connects Claude to Todoist, enabling full CRUD operations on tasks, projects, sections, comments, and labels.
README
⚠️ ARCHIVED: This project has been retired and is no longer maintained. Reason: Replaced by the official Todoist MCP plugin in the Claude connector registry. Retired: 2026-03-22
Todoist MCP Server
MCP server that connects Claude Desktop, Cowork, and Claude Code to your Todoist account. Full CRUD for tasks, projects, sections, comments, and labels.
Tools (25 total)
Tasks
todoist_list_tasks— List/filter tasks (by project, section, label, or Todoist filter syntax)todoist_get_task— Get a single task by IDtodoist_create_task— Create with due dates, priority, labels, recurring schedulestodoist_update_task— Update any task fieldtodoist_close_task— Mark complete (advances recurring tasks)todoist_reopen_task— Reopen a completed tasktodoist_delete_task— Permanently delete
Projects
todoist_list_projects/todoist_get_project/todoist_create_project/todoist_update_project/todoist_delete_project
Sections
todoist_list_sections/todoist_create_section/todoist_update_section/todoist_delete_section
Comments
todoist_list_comments/todoist_create_comment/todoist_update_comment/todoist_delete_comment
Labels
todoist_list_labels/todoist_create_label/todoist_update_label/todoist_delete_label
Prerequisites
- Todoist API Token — Get from: Todoist → Settings → Integrations → Developer
- Python 3.12+ (for local use) or Docker (for VPS deployment)
Quick Start — Local (stdio)
# Clone and install
git clone https://github.com/alobarquest/todoist-mcp.git
cd todoist-mcp
pip install -r requirements.txt
# Set your token
export TODOIST_API_TOKEN="your_token_here"
# Run (stdio mode for local clients)
python -m src.server
Local Development
Run Tests
pip install -r requirements.txt
python -m pytest tests/ -v
Lint
ruff check src/ tests/
Claude Desktop Configuration
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"todoist": {
"command": "python",
"args": ["-m", "src.server"],
"cwd": "/path/to/todoist-mcp",
"env": {
"TODOIST_API_TOKEN": "your_token_here"
}
}
}
}
Claude Code Configuration
Add to your project's .mcp.json or global ~/.claude/settings.json:
{
"mcpServers": {
"todoist": {
"command": "python",
"args": ["-m", "src.server"],
"cwd": "/path/to/todoist-mcp",
"env": {
"TODOIST_API_TOKEN": "your_token_here"
}
}
}
}
Remote Server (VPS via Coolify)
When deployed to the VPS, the server runs in streamable HTTP mode.
Connect Claude Desktop to the remote server:
{
"mcpServers": {
"todoist": {
"url": "https://todoist-mcp.devonwatkins.com/mcp/"
}
}
}
Environment variables for Coolify:
| Variable | Description |
|---|---|
TODOIST_API_TOKEN |
Your Todoist API token (from BWS) |
MCP_TRANSPORT |
streamable_http (set in Dockerfile) |
MCP_PORT |
8000 (set in Dockerfile) |
Deployment — Coolify/Hetzner VPS
This follows Flavor A from Devon's Infra Standards (lightweight, single container).
However, since we use GHCR image pulls (not source build), the CI/CD follows the Flavor B pattern:
- Push to
main→ GitHub Actions builds image → pushes toghcr.io/alobarquest/todoist-mcp - Coolify webhook triggers redeploy → pulls latest image
Coolify Setup
- Build type: Docker Image
- Image:
ghcr.io/alobarquest/todoist-mcp:latest - Port: 8000
- Domain:
todoist-mcp.devonwatkins.com - Health check:
http://127.0.0.1:8000/health/ready - Environment:
TODOIST_API_TOKENfrom BWS
GitHub Repo Secrets
COOLIFY_WEBHOOK_URL— Coolify deploy webhook for this resourceCOOLIFY_API_TOKEN— Coolify API bearer token
Health Checks
Two endpoints are available for health monitoring:
/health/live— Liveness check. Always returns 200 OK. Use to verify the server is running./health/ready— Readiness check. Returns 200 OK ifTODOIST_API_TOKENis set, 503 Service Unavailable otherwise. Use in Coolify health check configuration.
CI/CD
All commits to main are validated by GitHub Actions before deployment:
- Lint:
ruff check src/ tests/— code formatting and style - Tests:
python -m pytest tests/ -v— all tests must pass
The build-and-push job only runs after the validate job passes. PRs are also validated before merging.
Rate Limits
Todoist allows 1000 requests per user per 15 minutes. The server returns a clear error message if you hit this limit.
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。