Uptime Kuma MCP Server

This project provides a FastMCP (Message Control Protocol) server that acts as an intermediary to interact with an Uptime Kuma instance. It exposes a set of tools to manage monitors, tags, status pages, and retrieve heartbeat data from Uptime Kuma.

Features

The server exposes the following tools:

• Login to Uptime Kuma: (Internal function, used by all tools)
• add_monitors: Adds multiple HTTP monitors to Uptime Kuma.
• get_monitors: Retrieves a list of all monitors.
• delete_monitors: Deletes specified monitors by their IDs.
• pause_monitor: Pauses a specific monitor.
• resume_monitor: Resumes a specific monitor.
• get_monitor_beats: Retrieves heartbeats for a specific monitor.
• edit_monitor: Edits an existing monitor's attributes.
• add_monitor_tag: Adds a tag to a specific monitor.
• delete_monitor_tag: Deletes a tag from a specific monitor.
• get_status_page: Retrieves details for a specific status page.
• get_heartbeats: Retrieves heartbeats for a specific monitor within a time range, with pagination.
• get_tags: Retrieves all tags defined in Uptime Kuma.

How it Works

1. FastMCP Server: The core of the application is a FastMCP server instance. This server listens for incoming requests to execute predefined "tools".
2. Uptime Kuma API Integration: Each tool, when invoked, first attempts to log in to the configured Uptime Kuma instance using the uptime-kuma-api Python client library.
3. Asynchronous Operations: The server is built using asyncio for non-blocking I/O. Since the uptime-kuma-api library is synchronous, its calls are wrapped in loop.run_in_executor to prevent blocking the server's event loop.
4. Environment Configuration: Critical configuration details such as Uptime Kuma credentials and server binding information are managed via environment variables, loaded from a .env file.
5. Tool Definitions: Each exposed functionality (e.g., add_monitors, get_monitors) is defined as an asynchronous function decorated with @mcp.tool(). These tools handle the logic for interacting with the Uptime Kuma API and formatting the response.

Prerequisites

• Python 3.7+ (due to asyncio.get_running_loop())
• An Uptime Kuma instance.
• Access credentials (username, password) for the Uptime Kuma instance.

Setup

1. Clone the repository (if applicable):

   git clone <repository_url>
   cd uptime-kuma-mcp-server
   

2. Install dependencies: It's recommended to use a virtual environment.

   python -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt
   

Running the Server

The server can be run in two modes:

1. STDIO Mode (Default): This mode uses standard input/output for communication. It's typically used when the server is managed as a subprocess by another application.

   python main.py
   

   The server will start and log its activity to the console.

2. SSE (Server-Sent Events) Mode: This mode runs an HTTP server that communicates using Server-Sent Events. To run in this mode, you would need to modify the if __name__ == "__main__": block in main.py to call main_sse() instead of main_stdio().

   Example modification in main.py:

   if __name__ == "__main__":
       # logger.info("Starting MCP server in STDIO mode as per __main__ block...")
       # main_stdio()
       # logger.info("MCP server (stdio mode) has finished or exited.")
   
       logger.info("Starting MCP server in SSE mode...")
       main_sse() # Call main_sse() here
       logger.info("MCP server (SSE mode) has finished or exited.")
   

   Then run:

   python main.py
   

   The server will listen on the host and port specified by MCP_HOST and MCP_PORT environment variables (defaults to 0.0.0.0:8000).

Tool Details

All tools first attempt to log in to Uptime Kuma. If login fails, the tool execution will fail.

add_monitors(urls: list[str])

• Description: Adds multiple HTTP monitors to Uptime Kuma.
• Parameters:
  • urls (list of strings): A list of URLs to create monitors for. The monitor name is derived from the URL.
• Returns: A dictionary containing monitor_addition_results, kuma_url, total_requested, and successfully_added.

get_monitors()

• Description: Retrieves a list of all monitors from Uptime Kuma.
• Parameters: None.
• Returns: A dictionary containing monitors (trimmed list), total_count, and kuma_url.

delete_monitors(ids: list[int])

• Description: Deletes specified monitors by their IDs.
• Parameters:
  • ids (list of integers): A list of monitor IDs to delete.
• Returns: A dictionary containing monitor_deletion_results, total_requested, successfully_deleted, and kuma_url.

pause_monitor(monitor_id: int)

• Description: Pauses a specific monitor.
• Parameters:
  • monitor_id (integer): The ID of the monitor to pause.
• Returns: A dictionary indicating success or failure, along with API response.

resume_monitor(monitor_id: int)

• Description: Resumes a specific monitor.
• Parameters:
  • monitor_id (integer): The ID of the monitor to resume.
• Returns: A dictionary indicating success or failure, along with API response.

get_monitor_beats(monitor_id: int, hours: int = 1)

• Description: Retrieves heartbeats for a specific monitor within a given time range.
• Parameters:
  • monitor_id (integer): The ID of the monitor.
  • hours (integer, default: 1): Period in hours from now to retrieve beats for.
• Returns: A dictionary containing the beats, count, and status.

edit_monitor(monitor_id: int, options: dict)

• Description: Edits an existing monitor.
• Parameters:
  • monitor_id (integer): The ID of the monitor to edit.
  • options (dictionary): Monitor attributes to update (e.g., {"name": "New Name", "interval": 60}). Refer to Uptime Kuma API for available fields.
• Returns: A dictionary indicating success or failure, along with API response.

add_monitor_tag(monitor_id: int, tag_id: int, value: str)

• Description: Adds a tag to a specific monitor.
• Parameters:
  • monitor_id (integer): The ID of the monitor.
  • tag_id (integer): The ID of the tag to add.
  • value (string): The value for the tag.
• Returns: A dictionary indicating success or failure, along with API response.

delete_monitor_tag(monitor_id: int, tag_id: int)

• Description: Deletes a tag from a specific monitor.
• Parameters:
  • monitor_id (integer): The ID of the monitor.
  • tag_id (integer): The ID of the tag to delete.
• Returns: A dictionary indicating success or failure, along with API response.

get_status_page(slug: str)

• Description: Retrieves details for a specific status page by its slug.
• Parameters:
  • slug (string): The slug of the status page.
• Returns: A dictionary containing the status page data or an error message.

get_heartbeats(offset: int = 0, limit: int = 100)

• Description: Retrieves heartbeat monitor data with pagination.
  • Note: This tool fetches all heartbeats from Uptime Kuma first, then applies pagination on the server side. This might be inefficient for very large datasets.
• Parameters:
  • offset (integer, default: 0): Offset for pagination.
  • limit (integer, default: 100): Limit for number of heartbeats per page.
• Returns: A dictionary containing paginated data, total_available, count_returned, offset, and limit.

get_tags()

• Description: Retrieves all tags defined in Uptime Kuma.
• Parameters: None.
• Returns: A dictionary containing the list of tags (data) and their count.

Logging

The server uses Python's logging module. By default, logs are output to the console (standard output) with INFO level and include a timestamp, logger name, log level, and message.

Error Handling

• Configuration Errors: The server will fail to start (raising a ValueError) if essential Uptime Kuma environment variables (KUMA_URL, KUMA_USERNAME, KUMA_PASSWORD) are not set.
• API Errors: Errors during interaction with the Uptime Kuma API (e.g., login failure, invalid monitor ID) are logged. The tools typically return a dictionary with a "status" field indicating "error" or "failed" and an "error_message".
• General Exceptions: Uncaught exceptions within tool handlers are logged and re-raised, allowing the FastMCP framework to handle them appropriately (which might involve sending an error response to the client).