Hermes OS Cat — AI Command Center

A mobile-first command center for monitoring and controlling multi-agent AI systems in real time, with live telemetry, task orchestration, alerts, and remote agent interaction.

Comment

Hermes OS Cat is a mobile-first command center built to monitor, control, and understand a multi-agent AI system from anywhere.

I created it because traditional dashboards and terminal logs become difficult to follow once multiple agents, scheduled jobs, models, APIs, and background processes are running at the same time. I wanted one clear interface that could show what the entire system is doing, which agents are active, whether anything is stuck, and where intervention is required.

The application provides real-time visibility into agent activity, task execution, system health, alerts, model usage, scheduled workflows, and operational events. It allows me to remotely inspect the state of my AI infrastructure, follow live updates, identify failures, and interact with the system directly from a phone.

The interface was designed specifically for mobile use, allowing Hermes OS to function as a personal AI operations center rather than another desktop-only developer dashboard.

Key features include:

  • Real-time agent and task monitoring
  • Live telemetry and system health status
  • Multi-agent workflow visibility
  • Alerts for warnings, failures, and stuck processes
  • Scheduled job and automation tracking
  • Remote access from mobile devices
  • Model and provider activity monitoring
  • Operational history and debugging information
  • Mobile-first responsive interface
  • Real-time updates using Server-Sent Events

Hermes OS Cat was built with React, Vite, Node.js, TypeScript, REST APIs, Server-Sent Events, and custom integrations with my AI agent infrastructure.

The project was designed and developed independently from architecture to implementation. It combines frontend development, backend services, real-time communication, AI agent orchestration, observability, automation, and product design.

The current version serves as the operational interface for my personal Hermes OS environment. The next stage is expanding the live integrations, improving agent visualization, adding richer debugging tools, and connecting additional real system endpoints.

The long-term goal is to make complex AI systems understandable at a glance — without requiring users to constantly inspect terminals, logs, or multiple disconnected dashboards.

Built With

Share this project:

Updates