This is a practical guide for installing Hermes Agent on a VPS and using it for simple automations.

It is written for people who are comfortable copying commands, but do not want a deep Linux or DevOps explanation.

The setup in this guide is intentionally simple:

• One Ubuntu VPS
• One hosted AI model provider
• Hermes Agent installed on the VPS
• Telegram as the main interface
• Scheduled automations running in the background

Read them in this order:

1. 01-before-you-start.md
2. 02-create-and-connect-to-the-vps.md
3. 03-install-hermes-agent.md
4. 04-connect-telegram.md
5. 05-create-automations.md
6. 06-use-cases.md
7. 07-troubleshooting.md
8. 08-security-notes.md

Optional advanced guides:

• 09-bonus-busybee-cpu.md
• 10-bonus-rag-memory.md
• 11-bonus-local-inference-remote-gateway.md

There is also a longer publishing version:

• article-version.md

Hermes Agent is useful when you want an AI assistant that is always available, not tied to your laptop, and able to run scheduled jobs.

A few examples:

• Send you a daily AI news summary.
• Check if your websites are online.
• Monitor a VPS and warn you if disk or memory usage is high.
• Summarise GitHub activity every morning.
• Watch competitor websites for important changes.
• Create a weekly business report.
• Answer messages from Telegram while running on your VPS.

This guide does not try to explain every Hermes feature.

It focuses on the setup most people should start with:

• VPS
• Hermes
• Telegram
• Cron automations

Once that works, you can add more advanced channels, more tools, or custom integrations later.

• Website: https://hermes-agent.nousresearch.com/
• Documentation: https://hermes-agent.nousresearch.com/docs/
• Installation: https://hermes-agent.nousresearch.com/docs/getting-started/installation
• Quickstart: https://hermes-agent.nousresearch.com/docs/getting-started/quickstart
• Telegram guide: https://hermes-agent.nousresearch.com/docs/user-guide/messaging/telegram
• Cron automations: https://hermes-agent.nousresearch.com/docs/user-guide/features/cron

For most users, I would start with this:

• Ubuntu 24.04 LTS VPS
• 1 vCPU
• 2 GB RAM
• 20 GB disk
• Telegram bot access
• Hosted model provider, not a local model

You do not need a GPU for this setup.

The VPS runs Hermes. The AI model runs through a provider API.