Heartbeat

Analyze cron-state.json, memory logs, and run statuses to surface skill failures.

Updated Jun 2, 2026

Search Tags:#automation #monitoring #memory #heartbeat #ai-ops #skill-health

Ready to run

npx skills add https://github.com/Atrium-Hermes/atrium-lighthouse --skill heartbeat-atrium-hermes

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Heartbeat
Source: https://github.com/Atrium-Hermes/atrium-lighthouse/tree/main/skills/heartbeat
Command: npx skills add https://github.com/Atrium-Hermes/atrium-lighthouse --skill heartbeat-atrium-hermes

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Proactive ambient checks surface failures, stalls, and degradation in a fleet of running skills, helping operators stay ahead of issues.

Core Features & Use Cases

Continuous heartbeat: surfaces failed or stuck skills, degraded performance, and memory anomalies for quick remediation.
Fleet-wide health: analyzes memory/cron-state.json, memory logs, aeon.yml, and last run statuses to highlight issues.
Actionable guidance: provides recommended next steps and health snapshots to stabilize operations.

Quick Start

Run the heartbeat check to generate a health snapshot across all enabled skills.

Frequently Asked Questions about Heartbeat

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I monitor memory and skill health to surface failures across running tasks?▼

You can monitor memory and skill health by running a heartbeat check that analyzes cron-state.json, last run statuses, and memory logs to surface failures, stalls, and degradation. This generates a structured health snapshot for quick remediation.

What is a heartbeat check and how does it detect skill degradation?▼

A heartbeat check is a proactive ambient monitoring process that reads memory files like cron-state.json and MEMORY.md to detect skill degradation. It identifies failed or stuck skills by analyzing last run statuses and memory anomalies.

Can I use runtime auditing to detect stuck or failed skills from memory logs?▼

Yes, runtime auditing detects stuck or failed skills by reading memory logs and cron-state.json. The heartbeat assessment analyzes these memory files to highlight stalls and provides actionable guidance to stabilize operations.

Do I need aeon.yml and cron-state.json files to generate a fleet-wide health snapshot?▼

Yes, generating a fleet-wide health snapshot requires reading aeon.yml and memory/cron-state.json. These files provide the deterministic state data needed to surface issues and assess the overall health of running skills.

How do I run a daily health check to spot memory anomalies and performance degradation?▼

Run a heartbeat check during daily health checks to spot memory anomalies and performance degradation. It reads memory logs and optional token reports to surface issues, providing a structured assessment with recommended next steps.