Heartbeat Silence Check
Install when your agent stack writes heartbeats to workspace/system_memory and you need pings if collection goes quiet too long.
Overview
Heartbeat Silence Check is an Operate-phase MCP server that alerts you when workspace/system_memory/heartbeats.json has not been updated within the silence threshold (default 6 hours).
What is this MCP server?
- Compares heartbeats.json modification time against a configurable silence threshold
- Default silence threshold of six hours before firing
- Voice alerts and Telegram pings at high priority
- Targets workspace/system_memory/heartbeats.json as the canonical signal file
- Designed as a counter-measure after multi-day heartbeat collector outages
- Default silence threshold: 6 hours since heartbeats.json mtime
- Canonical path: workspace/system_memory/heartbeats.json
- Server version 0.1.0, MIT license
What problem does it solve?
Autonomous heartbeat collectors can stop writing silently for days while you assume the agent fleet is still healthy.
Who is it for?
Solo operators running OpenClaw or DailyAI OS with heartbeats.json and Telegram or voice notification hooks already configured.
Skip if: Static repos with no scheduled agents or teams that monitor uptime only through hosted SaaS APM.
What do I get? / Deliverables
You receive voice and Telegram high-priority alerts as soon as heartbeat freshness fails, so you revive collectors before extended downtime.
- Silence detection when heartbeats.json exceeds the threshold age
- High-priority voice alert when configured
- High-priority Telegram ping when configured
Recommended MCP Servers
Journey fit
Operate is the right shelf because the server exists to detect production silence when autonomous collectors stop updating system memory. Monitoring matches mtime checks on heartbeats.json and outbound high-priority alerts when the default six-hour window is breached.
How it compares
MCP freshness watchdog for one JSON heartbeat file, not a full metrics platform or log aggregator.
Common Questions / FAQ
Who is heartbeat-silence-check for?
Builders operating local agent workspaces that persist heartbeats under system_memory and need human-visible alerts when updates stop.
When should I use heartbeat-silence-check?
Use it in Operate whenever a heartbeat-collector should run continuously and silence longer than about six hours means something archived or crashed.
How do I add heartbeat-silence-check to my agent?
Add the io.github.dailyaiagents-cpu/heartbeat-silence-check MCP server to your client config, ensure heartbeats.json exists on the watched path, and connect Telegram or voice channels your stack supports.