Radicale CalDAV MCP Server
Enables LLM agents to read, schedule, update, and delete calendar events and tasks on a self-hosted CalDAV server, and to query availability for free/busy slots.
README
Radicale CalDAV MCP Server
A Model Context Protocol (MCP) server built with FastMCP / the official MCP Python SDK that bridges LLM agents (such as Nous Hermes, Claude, OpenAI) to a self-hosted Radicale CalDAV calendar and task manager.
Features
📅 Calendar Events (VEVENT)
get_events: Query events within date/time ranges with optional keyword search across summary, description, and location.get_free_busy_slots: Calculate open, unbooked time slots during configurable business hours (e.g. 9:00 AM - 5:00 PM) in any IANA timezone.create_event: Schedule new events with automatic ISO validation, attendee invitations, and overlap/conflict detection.update_event: Modify existing events by UID without overwriting untouched fields.delete_event: Remove events by UID.
📝 Tasks & To-Dos (VTODO)
list_todos: Fetch pending or completed tasks filtered by status (NEEDS-ACTION,COMPLETED,IN-PROCESS,CANCELLED,ALL).create_todo: Create tasks with due dates, priority levels (1-9), and descriptions.complete_todo: Mark tasks as completed with automatic completion timestamps.
🚀 Transports & Deployment
- STDIO Transport: Default transport for local execution (e.g. Claude Desktop, local agents).
- SSE (Server-Sent Events) Transport: Over HTTP for remote, always-on deployments (e.g. Coolify, Docker).
Configuration
Configure the server via environment variables or a .env file:
| Variable | Default | Description |
|---|---|---|
RADICALE_BASE_URL |
http://localhost:5232 |
Base URL to Radicale (e.g. https://radicale.dubblu.app) |
RADICALE_USERNAME |
mvela581 |
CalDAV username |
RADICALE_PASSWORD |
"" |
CalDAV password |
RADICALE_CALENDAR_SLUG |
seuspeptides |
Target calendar slug / collection name |
DEFAULT_TIMEZONE |
America/New_York |
Default IANA timezone |
MCP_TRANSPORT |
stdio |
Transport protocol: stdio or sse |
MCP_HOST |
0.0.0.0 |
Host interface for SSE mode |
MCP_PORT |
8000 |
Port number for SSE mode |
Local Quickstart
1. Installation
# Clone and enter directory
cd radicale-mcp-server
# Create virtual environment
python3 -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install -r requirements.txt
2. Configure Environment
Copy .env.example to .env and set your credentials:
cp .env.example .env
3. Run Locally
STDIO Mode (Local Pipe)
python server.py --stdio
SSE Mode (HTTP Server)
python server.py --sse --host 0.0.0.0 --port 8000
4. Run Test Client
# Test via STDIO
python test_client.py --mode stdio
# Test via SSE (when server is running)
python test_client.py --mode sse --sse-url http://localhost:8000/sse
5. Run Automated Tests
pytest -v
Coolify Deployment
Deploy the MCP server on Coolify with zero configuration:
- Create a new service in Coolify: Docker Compose or Dockerfile.
- Point it to this repository or paste the
docker-compose.yml. - Set the following environment variables in Coolify UI:
RADICALE_BASE_URL=https://radicale.dubblu.appRADICALE_USERNAME=mvela581RADICALE_PASSWORD=your_passwordRADICALE_CALENDAR_SLUG=seuspeptidesDEFAULT_TIMEZONE=America/New_YorkMCP_TRANSPORT=sseMCP_PORT=8000
- Expose port
8000via Coolify reverse proxy with your domain (e.g.https://mcp-calendar.yourdomain.com). - Your SSE endpoint will be accessible at:
https://mcp-calendar.yourdomain.com/sse
LLM Agent Integration (Nous Hermes / Claude Desktop)
Claude Desktop Configuration
Add the following to your claude_desktop_config.json:
Local STDIO:
{
"mcpServers": {
"radicale_calendar": {
"command": "/path/to/radicale-mcp-server/.venv/bin/python",
"args": ["/path/to/radicale-mcp-server/server.py", "--stdio"],
"env": {
"RADICALE_BASE_URL": "https://radicale.dubblu.app",
"RADICALE_USERNAME": "mvela581",
"RADICALE_PASSWORD": "your_password",
"RADICALE_CALENDAR_SLUG": "seuspeptides",
"DEFAULT_TIMEZONE": "America/New_York"
}
}
}
}
Remote SSE (Coolify):
{
"mcpServers": {
"radicale_calendar": {
"url": "https://mcp-calendar.yourdomain.com/sse"
}
}
}
Nous Hermes Agent Integration
Tool definitions are auto-generated from docstrings with type specifications. Sample prompt usage:
{
"name": "get_free_busy_slots",
"arguments": {
"start_date": "2026-08-26",
"end_date": "2026-08-28",
"slot_duration_minutes": 30,
"working_hours_start": 9,
"working_hours_end": 17,
"timezone": "America/New_York"
}
}
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。
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。
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。
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。