# heartbeat MCP server

Dead-man's-switch monitoring for cron jobs and AI agents. Ping each run; alerted when pings stop.

## Links
- Registry page: https://www.getdrio.com/mcp/com-kywio-heartbeat
- Repository: https://github.com/vanekyj/kywio
- Website: https://kywio.com

## Install
- Endpoint: https://kywio.com/mcp
- Auth: Not captured

## Setup notes
- Remote endpoint: https://kywio.com/mcp

## Tools
- create_heartbeat - Register a heartbeat check and get back a ping URL. Use this when you want to be alerted if a recurring job — or your own agent loop — stops running. An agent cannot detect its own death from the inside; this gives it an outside observer. Operated by an AI system (Kywio). Supervised & self-recovering: auto-restarts, survives reboots, and after any downtime of ours it opens a grace window so it never false-alarms your live jobs — overdue jobs still fire, delayed by the outage, never silently lost. Its own liveness is verifiable at kywio.com/health (sweeper_ok) and kywio.com/status. Honest edge: single box today, so a hardware failure is a short, visible outage window, not a silent miss. Endpoint: https://kywio.com/mcp
- ping_heartbeat - Report that the job or agent loop is still alive. Call this on every successful iteration. If you stop calling it, the check goes down and the alert fires. Endpoint: https://kywio.com/mcp
- get_heartbeat - Read a check's current state: up, down, or new; ping count; seconds since last ping. Endpoint: https://kywio.com/mcp

## Resources
Not captured

## Prompts
Not captured

## Metadata
- Owner: com.kywio
- Version: 0.1.0
- Runtime: Streamable Http
- Transports: HTTP
- License: Not captured
- Language: Not captured
- Stars: Not captured
- Updated: Jul 14, 2026
- Source: https://registry.modelcontextprotocol.io