Skip to main content
Operations covers everything after installation — keeping your Comis installation running smoothly, understanding what it is doing, and knowing how to fix things when they go wrong. Whether you just finished the Quickstart or are managing a production deployment, this section is your guide to day-to-day Comis operations.

​
What Operations Covers

Operations is organized into three groups:

​
Management

These pages explain the systems that keep Comis running and how to understand what is happening inside your installation.
  • Daemon — The background process that powers everything. How it starts, runs, and shuts down.
  • Data Directory — Every file Comis keeps in ~/.comis/, what each one does, and what is safe to delete.
  • Logging — Where logs go, what the levels mean, and how to view them.
  • Monitoring — Automatic health checks that watch disk space, CPU, memory, and more.
  • Observability — Token tracking, cost estimation, latency, and delivery tracing.
  • Scheduler — Cron jobs, heartbeat checks, quiet hours, and task extraction.
  • Web UI — Setting up the browser-based dashboard for managing Comis.

​
Deployment

These pages cover how to run Comis in production with different process managers and reverse proxies.
  • systemd — Running Comis as a Linux system service with automatic restarts and watchdog monitoring.
  • Docker — Production-ready container setup with health checks and volume management.
  • Docker Hub — Pull official images from Docker Hub and the multi-arch tag scheme.
  • pm2 — Using the pm2 process manager for easy start/stop/restart workflows.
  • Reverse Proxy — Putting Comis behind Nginx or Caddy with TLS termination.

​
Help

When something goes wrong or you have questions, start here.
  • Troubleshooting — Common problems and how to solve them, with exact error messages and step-by-step fixes.
  • FAQ — Answers to frequently asked questions about running Comis.

​
Quick Health Check

If you just want to confirm that Comis is running and healthy, follow these three steps:
1

Check if the daemon is running

pm2 status comis
You should see a table with comis showing status online:
┌─────┬────────┬──────┬───────┬────────┬─────────┬────────┐
│ id  │ name   │ mode │ pid   │ status │ restart │ uptime │
├─────┼────────┼──────┼───────┼────────┼─────────┼────────┤
│ 0   │ comis  │ fork │ 12345 │ online │ 0       │ 2h     │
└─────┴────────┴──────┴───────┴────────┴─────────┴────────┘
2

Check the health endpoint

The daemon exposes a health endpoint you can check from the command line:
curl http://localhost:4766/health
Expected response:
{"status":"ok","timestamp":"2026-03-12T10:00:00.000Z"}
If you get a connection error, the daemon is not running or the gateway is not enabled.
3

View recent logs

pm2 logs comis --lines 20 --nostream
Look for "Comis daemon started" in the output — that confirms the daemon initialized successfully. If you see FATAL: errors, check the Troubleshooting page.
If you just installed Comis, start with the Daemon page to understand how the background process works. Then visit Logging to learn how to read what Comis is telling you.

​
All Operations Pages

Daemon

How the background process starts, runs, and shuts down.

Data Directory

Every file in ~/.comis/ and what it does.

Logging

Understanding and viewing Comis logs.

Monitoring

Health monitoring and system alerts.

Observability

Token tracking, cost estimation, and delivery tracing.

Scheduler

Cron jobs, heartbeat checks, and task extraction.

Web UI

Setting up the browser-based dashboard.

systemd

Running Comis as a Linux system service.

Docker

Production-ready container setup.

Docker Hub

Pull official comisai/comis images.

pm2

Easy process management with pm2.

Reverse Proxy

TLS termination with Nginx or Caddy.

Troubleshooting

Common problems and step-by-step fixes.

FAQ

Frequently asked questions about running Comis.