monitors

Manage long-running background monitors for orchestration state and external systems.

Updated Mar 9, 2026
One-click install
npx skills add https://github.com/ether-moon/agent-atelier --skill monitors
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: monitors
Source: https://github.com/ether-moon/agent-atelier/tree/main/plugins/agent-atelier/skills/monitors
Command: npx skills add https://github.com/ether-moon/agent-atelier --skill monitors

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Monitors provide a structured, long-running observation layer for an orchestration loop. They enable automated health checks, event polling, and lifecycle management of background tasks, ensuring the system stays responsive and debuggable without blocking main work.

Core Features & Use Cases

  • Spawn and manage multiple always-on monitors (heartbeat, gate, events, divergence) to observe state and external signals.
  • Poll for events, assess health, and report status, so the orchestrator can react quickly to issues.
  • Start on loop initiation, support on-demand CI monitors, and gracefully stop during cleanup.

Quick Start

Start the monitors at loop begin with the monitors spawn command.

Frequently Asked Questions about monitors

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

FAQPage Schema
How do I monitor background processes during orchestration?

Background processes are monitored by spawning always-on monitors that observe orchestration state and external systems, running health checks and reporting status back to the orchestrator.

What is the best way to track CI status in an automated loop?

CI status is tracked by spawning on-demand CI monitors that poll external systems and report results back to the orchestrator without blocking main workflow execution.

How do I start and stop background monitors for event polling?

Background monitors start at loop initialization using the spawn command and gracefully stop during cleanup, handling event polling and lifecycle management throughout the loop.

Can I run multiple health checks simultaneously without blocking the main task?

Multiple health checks run simultaneously through always-on monitors like heartbeat, gate, events, and divergence types, keeping the system responsive without blocking main work.

Do I need the agent-atelier skill to spawn orchestration monitors?

The monitors skill relies on the agent-atelier skill's scripts and references to spawn, monitor, and report back to the Orchestrator, making it a required dependency for execution.