mcp-deadmansnitch

This MCP server provides a unified tool named snitch to manage Dead Man's Snitch monitors via the MCP interface. It supports listing snitches, creating and updating monitors, checking in, pausing and unpausing monitors, and tagging for organization. Use the snitch action with appropriate parameters to perform common tasks, such as listing all snitches, creating a daily monitor, or sending a check-in for a specific token. The action set is designed to minimize context switching for LLMs by consolidating all Dead Man's Snitch operations behind a single tool endpoint.

Prerequisites: a supported Python environment and network access. Ensure you have Python 3.8+ and access to install Python packages. Steps:\n1) Install Python and pip if not already installed.\n2) Install the MCP package runner (uvx) which runs MCP servers:\n - python -m pip install uv\n3) Install the MCP server package for Dead Man's Snitch:\n - python -m pip install mcp-deadmansnitch\n4) Run the MCP server via the recommended uvx command:\n - uvx mcp-deadmansnitch\n5) Configure your MCP client to point at the server using the provided configuration example (see README).\n6) Ensure your Dead Man's Snitch API key is available to the client by setting DEADMANSNITCH_API_KEY in environment variables or via your client configuration.

Environment variables: Set DEADMANSNITCH_API_KEY to your Dead Man's Snitch API key. The MCP server exposes a single tool named snitch with actions like list, get, create, update, delete, pause, unpause, check_in, add_tags, and remove_tag. Valid intervals for snitches include 15_minute, hourly, daily, weekly, monthly, and valid alert_types include basic and smart. If running via Docker or Nix, adapt the command accordingly and ensure the API key is exposed to the container or environment. If you encounter authentication errors, double-check the API key and that it has the necessary permissions for the requested operations.