Hermes

How to Install Hermes Agent: macOS, Windows & Linux Server (+ Skills Setup)

· 7 min read · YayaAgent Team

Hermes Agent is an open-source autonomous AI agent built by Nous Research. What sets it apart: a built-in learning loop that creates skills from your sessions, improves them over time, and remembers context across conversations. This guide gets you up and running in minutes.

macOS · Windows · Linux Free & open source (MIT)
Part 1 — Install Hermes Agent
Heads up: The desktop app (macOS & Windows) is as easy as installing any app — download, double-click, done. However, Hermes only runs while your computer is on and the app is open. If you want your agent available 24/7, consider a Linux server deployment instead — see the Linux Server tab below.
Download the Desktop App

Go to the official site and click Download for macOS:

Download: hermes-agent.nousresearch.com
Open the installer and follow the steps

Double-click the downloaded file and follow the on-screen prompts. The installer sets up both the desktop app and the command-line tool automatically — no manual configuration needed.

Connect your AI model

On first launch, the setup wizard asks which AI model to use. The easiest option is Nous Portal — one subscription covers 300+ models plus web search and image generation:

Prefer your own API key? Choose Anthropic, OpenAI, or OpenRouter during setup. You can always switch later with hermes model in the terminal.
Start chatting

Type your first message in the app. You can also connect Telegram or Discord so you can talk to your agent from your phone:

# In the built-in terminal inside the app
hermes gateway setup
Download the Desktop App

Go to the official site and click Download for Windows:

Download: hermes-agent.nousresearch.com
Run the installer

Double-click the downloaded file. Windows may show a SmartScreen warning — click More info → Run anyway. This is a known issue (the installer isn't code-signed yet) acknowledged in the official docs — it's safe.

No prerequisites needed. The installer bundles everything, including its own portable Git. It won't interfere with anything already on your system.
Connect your AI model & start chatting

Follow the setup wizard on first launch to connect your AI model, then start your first conversation directly in the app.

Antivirus warning: Some tools (Bitdefender, Windows Defender) may flag uv.exe in the Hermes folder. This is a false positive — it's a safe Rust-based package manager. Add an exception if prompted.
Best for 24/7 availability. Run Hermes on a VPS or home server and talk to it via Telegram from anywhere — even when your laptop is off.
Install prerequisites (Debian/Ubuntu)
sudo apt update && sudo apt install -y git curl xz-utils

That's all you need to pre-install. Python, Node.js, and other dependencies are handled by the Hermes installer automatically.

Run the one-line installer
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash

Takes 2–3 minutes. When done, reload your shell:

source ~/.bashrc
Headless server? Add --skip-browser to skip the Chromium download if you don't need web browsing:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash -s -- --skip-browser
Connect your AI model
hermes setup # Choose: Nous Portal, Anthropic, OpenAI, OpenRouter...
Connect Telegram (recommended for servers)

No GUI on a server — Telegram lets you chat with your agent from your phone while it runs 24/7:

hermes gateway setup # Choose Telegram → follow the bot setup steps
Keep Hermes running after you close SSH
sudo tee /etc/systemd/system/hermes.service <<EOF
[Unit]
Description=Hermes Agent
After=network.target

[Service]
User=$USER
ExecStart=$(which hermes) gateway
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

sudo systemctl enable --now hermes

Hermes will now start automatically on every reboot.

✅ Quick sanity check

  • Hermes opens and responds to your first message
  • You've connected at least one AI model provider
  • (Optional) Your Telegram bot replies when you message it
Part 2 — Installing Skills

Skills are what make Hermes genuinely useful beyond basic chat. Each skill is a SKILL.md file — plain Markdown that teaches your agent how to handle a specific type of task. Hermes also creates skills automatically after complex sessions, so it genuinely gets better the more you use it.

Good news: Hermes ships with a built-in skill catalog. You don't need to install anything extra to get started — but community skills unlock much more.
🔍 web-research
Search the web and summarize results on demand
📝 blog-writer
Long-form SEO writing with keyword control
💻 github-pr-workflow
Create PRs and review code changes via chat
📅 calendar
Read and manage your calendar from any channel
🎨 excalidraw
Generate diagrams from natural language
📊 sql-toolkit
Query databases via natural language
Search for a skill
hermes skills search "calendar"
hermes skills search "github"

You can also browse the community registry at agentskills.io.

Install a skill
hermes skills install skill-name

Available immediately — no restart needed. Each skill also becomes a slash command in chat:

/calendar what's on my schedule tomorrow?
Security: All Skills Hub installs go through an automated security scanner. Still, inspect any third-party skill before use with hermes skills inspect skill-name.
List, update, and check skills
hermes skills list # see installed skills
hermes skills update # update all skills
hermes skills check # check for upstream changes
Where skills live on disk
~/.hermes/skills/ # global (all projects)
./skills/ # project-level (current folder)
Self-improving: After completing a complex task, Hermes automatically drafts a skill for it and asks if you want to save it — no manual action needed.

🚀 You're all set

Run hermes update regularly to stay current. Explore community skills at agentskills.io. The more you use Hermes, the more it learns — give it real tasks and let it build its own skill library over time.