Spin up and harden your Hostinger KVM-2 VPS in under 10 minutes.

1. Deploy the VPS

• Log into Hostinger hPanel → VPS → Deploy
• Choose KVM-2 plan (Ubuntu 24.04 LTS)
• Set a strong root password, save the SSH credentials
• Wait ~2 minutes for provisioning, note the IP address

2. Initial SSH & Hardening

• SSH in: ssh root@<your-vps-ip>
• Run updates: apt update && apt upgrade -y
• Create a non-root user: adduser claw and usermod -aG sudo claw
• Set up SSH key auth for claw user, disable root login & password auth in /etc/ssh/sshd_config
• Restart SSH: systemctl restart ssh
• Enable UFW: ufw allow OpenSSH && ufw enable

3. Install Node.js & Dependencies

• Install Node.js 22: curl -fsSL https://deb.nodesource.com/setup_22.x | bash - && apt install -y nodejs
• Verify: node -v && npm -v
• Install PM2 for process management: npm install -g pm2
• Install Git: apt install -y git

4. Deploy Open Claw

• Clone: git clone https://github.com/openclaw/openclaw.git /opt/openclaw
• Install: cd /opt/openclaw && npm install
• Configure: edit config.yaml (set agents, tools, runtime)
• Start with PM2: pm2 start openclaw.js --name openclaw
• Save PM2 process list: pm2 save && pm2 startup

5. Open Ports for Webhook (if needed)

• ufw allow 80/tcp (HTTP) or 443/tcp (HTTPS with reverse proxy)
• Optionally set up Nginx + Let's Encrypt for TLS termination

OpenRouter gives Open Claw access to dozens of LLMs through a single API — from GPT-4o to Claude 3.5 Sonnet to DeepSeek V3 — with pay-per-token pricing.

1. Get Your API Key

• Go to openrouter.ai/keys
• Create a new key, copy it immediately
• Add $5-10 credit to your OpenRouter account

2. Configure Open Claw

• In your Open Claw config, set the LLM provider to OpenRouter:

llm:
  provider: openrouter
  apiKey: "sk-or-v1-xxxxxxxxxxxxxxxx"
  model: "openai/gpt-4o"
  temperature: 0.7
  maxTokens: 4096

3. Recommended Models

4. Fallback & Routing

• Configure model fallbacks so if GPT-4o is down, it falls back to Claude:

fallbacks:
  - "anthropic/claude-3.5-sonnet"
  - "deepseek/deepseek-chat"

Connect Open Claw to Telegram so you can chat with your AI agent from your phone, desktop, or anywhere.

1. Create a Telegram Bot

• Open Telegram and search for @BotFather
• Send: /newbot
• Follow prompts: set bot name (e.g. ClawBot) and username (e.g. claw_agent_bot)
• Copy the API token BotFather gives you (starts with 7xxxxxx:AAH...xxx)
• Optional: set a profile photo via /setuserpic

2. Configure Telegram in Open Claw

• In your config.yaml, add the Telegram channel:

channels:
  telegram:
    enabled: true
    botToken: "7xxxxxx:AAH...xxx"
    allowedUserIds:
      - "<your-telegram-id>"

3. Polling vs Webhook

• Polling (simpler, works behind NAT): Open Claw polls Telegram every few seconds — no webhook setup needed. Works out of the box.
• Webhook (faster, recommended for production): requires a public HTTPS endpoint. Set up Nginx reverse proxy + Let's Encrypt on your VPS, then tell Telegram to POST updates to https://your-domain.com/telegram-webhook

4. Test It

• Open your bot on Telegram, send /start
• If it replies, you're connected. Try asking it questions or giving it tasks.
• Monitor logs: pm2 logs openclaw

Core CLI commands for operating OpenClaw on your Hostinger VPS and Mac Mini. These evolve quickly — this covers the essentials.

Agent Commands

# Chat with an agent
openclaw agent --agent main "Hello"

# Force local execution (bypass Gateway)
openclaw agent --local --agent main "What time is it?"

# Send to a specific session
openclaw agent --session-id abc123 "Continue"

Agent Management

# List all agents
openclaw agents list

# Add a new agent
openclaw agents add sales-agent

# Modify agent identity (name, emoji, avatar, color, description)
openclaw agents set-identity sales-agent

# Remove an agent
openclaw agents remove sales-agent

Gateway Commands

The Gateway is the daemon that Telegram, Discord, WhatsApp and other channels talk to.

# Start the background service
openclaw gateway start

# Restart (most-used command after config changes)
openclaw gateway restart

# Stop the daemon
openclaw gateway stop

# Check if it's running
openclaw gateway status

Configuration Commands

# View full configuration
openclaw config get

# View specific setting
openclaw config get models

# Set a value
openclaw config set agents.defaults.models.chat deepseek/deepseek-v4

# Export config (useful before upgrades)
openclaw config export

Model Commands

# List available models and providers
openclaw models list

# Test a model directly
openclaw agent --model deepseek/deepseek-v4 "Hello"

Session Commands

# List active and historical sessions
openclaw sessions list

# Cleanup old sessions (frees disk)
openclaw sessions cleanup

# Cleanup across all agents
openclaw sessions cleanup --all-agents

Channel Commands

# List configured channels
openclaw channels list

# Check channel connectivity
openclaw channels status

# Deep probe
openclaw channels status --probe

Cron / Automation Commands

# List scheduled jobs
openclaw cron list

# Run a job manually (useful for testing)
openclaw cron run daily-summary

Logging Commands

# Follow logs live (equivalent to tail -f)
openclaw logs --follow

# View recent logs
openclaw logs

Troubleshooting Commands

# Run diagnostics (first thing when something breaks)
openclaw doctor

# Check version
openclaw --version

# List all commands
openclaw --help

📋 Daily Ops Cheat Sheet

Commands you'll use most often on your VPS deployment:

openclaw gateway restart
gateway restart

openclaw logs --follow
openclaw doctor
openclaw channels status --probe
openclaw config get
openclaw sessions list
openclaw cron list

These seven commands solve about 90% of operational issues. For Hostinger-specific ops — restart after VPS reboot, check Telegram connectivity, switch models, view active agents, or set up systemd for auto-start on reboot — just ask.

Provision and secure your Hostinger KVM-2 VPS for Hermes in minutes.

1. Deploy the VPS

• Hostinger hPanel → VPS → Deploy → KVM-2 plan
• Select Ubuntu 24.04 LTS
• Set root password, SSH in once provisioned

2. System Setup & Hardening

• ssh root@<vps-ip>
• apt update && apt upgrade -y
• Create user: adduser hermes && usermod -aG sudo hermes
• Set up SSH keys, disable root login + password auth
• Firewall: ufw default deny incoming && ufw allow OpenSSH && ufw enable
• Set timezone: timedatectl set-timezone UTC

3. Install Hermes

• apt install -y git curl
• curl -fsSL https://deb.nodesource.com/setup_22.x | bash - && apt install -y nodejs
• npm install -g pm2
• git clone https://github.com/your-org/hermes.git /opt/hermes
• cd /opt/hermes && npm install
• Copy config template: cp .env.example .env
• Edit .env with your settings
• pm2 start index.js --name hermes && pm2 save && pm2 startup

OpenRouter gives Hermes access to the best LLMs at competitive per-token rates — no long-term commitments or API quotas.

1. Get Your OpenRouter Key

• Visit openrouter.ai/keys
• Create an API key, add initial credit

2. Configure Hermes

• In your .env file:

LLM_PROVIDER=openrouter
OPENROUTER_API_KEY="sk-or-v1-xxxxxxxxxxxxxxxx"
LLM_MODEL="openai/gpt-4o"
LLM_TEMPERATURE=0.7
LLM_MAX_TOKENS=4096

3. Cost-Efficient Model Strategy

• Primary: GPT-4o for complex reasoning and tool use
• Simple tasks: DeepSeek V3 or Qwen 2.5 (80-90% cheaper)
• Vision: Claude 3.5 Sonnet for image understanding
• OpenRouter lets you switch models per-request via the model parameter

Wire up Telegram as Hermes' chat interface so you can talk to your agent from any device.

1. Create the Bot

• Message @BotFather on Telegram
• /newbot → name it (e.g. Hermes Agent) → username (e.g. hermes_agent_bot)
• Save the API token

2. Configure in Hermes

• Add to .env:

TELEGRAM_ENABLED=true
TELEGRAM_BOT_TOKEN="7xxxxxx:AAH...xxx"
TELEGRAM_ALLOWED_USERS="<your-user-id>"

3. Run & Test

• Restart Hermes: pm2 restart hermes
• Check logs: pm2 logs hermes
• Send /start to your bot on Telegram
• Try commands like /tools or just type a question

Core Hermes CLI commands for daily operation on your VPS and Mac Mini.

Chat Commands

# Start interactive Hermes CLI
hermes
# or
hermes chat

# One-shot prompt
hermes chat -q "Summarize this folder"

# Clean output for scripts/cron
hermes -z "What is the capital of France?"

Setup & Model Commands

# Full setup wizard
hermes setup

# Choose or change model provider
hermes model

# Switch model inside chat
/model claude-sonnet-4
/model openrouter:anthropic/claude-sonnet-4
/model claude-sonnet-4 --global

Gateway Commands

Connects Hermes to Telegram, Discord, Slack, WhatsApp, Signal, etc.

# Interactive setup for messaging platforms
hermes gateway setup

# Run gateway in foreground (Docker, tmux, testing)
hermes gateway run

# Start/stop/restart as background service
hermes gateway start
hermes gateway stop
hermes gateway restart

# Check gateway status
hermes gateway status

# Install as systemd (Linux) or launchd (macOS)
hermes gateway install

Status & Troubleshooting

# Health check (first thing when something breaks)
hermes doctor

# Overall status
hermes status
hermes status --all
hermes status --deep

# Debug dump (redacted, copy-pasteable for support)
hermes dump

# View logs
hermes logs

Sessions

# List sessions
hermes sessions list

# Browse sessions
hermes sessions browse

# Resume last session
hermes --continue

# Resume a specific session
hermes --resume <session-id-or-title>

# Delete / prune sessions
hermes sessions delete <session-id>
hermes sessions prune

Profiles

# List profiles
hermes profile list

# Use a specific profile
hermes profile use work

# Create a profile (optionally clone existing)
hermes profile create work --clone

# Show profile details
hermes profile show work

Tools, Skills & Automation

# Configure tools (enable/disable per platform)
hermes tools

# Cron / scheduled jobs
hermes cron list
hermes cron create
hermes cron pause
hermes cron resume
hermes cron run <job>
hermes cron remove <job>

Maintenance

# Version
hermes --version
hermes version

# Update Hermes
hermes update

# Check for update only
hermes update --check

# Uninstall
hermes uninstall
hermes uninstall --full

📋 Daily Ops Cheat Sheet

Most-used commands on your VPS:

hermes doctor
hermes status --deep
hermes model
hermes gateway status
hermes gateway restart
hermes logs
hermes sessions list
hermes --continue

Key operational pair for your Hostinger VPS: hermes gateway status and hermes gateway restart.