Skip to main content
🎓 Claude Code Masterclass Learn AI-assisted development on Udemy — plus the companion book on Leanpub & Amazon. Start Learning
Troubleshooting diagram for Hermes Agent on a 1GB RAM VPS: PATH, swap, and gateway command fixes
AI

Hermes Agent on a 1GB VPS: Fixing the Common Gotchas

Hermes Agent install issues on a free-tier 1GB VPS: 'command not found' after install, Playwright/Chromium failures, and gateway subcommand mistakes.

LB
Luca Berton
· 3 min read

The Problem

You’re setting up Hermes Agent on a small, resource-constrained VPS — often a free-tier instance with 1 GB of RAM, like Oracle Cloud’s VM.Standard.E2.1.Micro shape — and hit one of a handful of recurring snags. None of these are exotic; they show up because a 1 GB box has very little headroom and because it’s easy to run the wrong command order under the assumption it’ll “just work.”

This covers the three most common ones, in the order you’re likely to hit them.

Issue 1: “hermes: command not found” Right After Install

Symptom: The installer finishes cleanly, prints ✓ hermes command ready, but running hermes — in the same shell, or a new one — returns:

Command 'hermes' not found, did you mean:
  command 'heroes' from deb heroes (0.21-18)

Cause: The install script puts the hermes launcher in ~/.local/bin and appends a PATH update to ~/.bashrc. A shell session that was already running before the install finished has no way to know about that change until it re-reads its startup files.

Fix:

source ~/.bashrc
export PATH="$HOME/.local/bin:$PATH"
which hermes

If which hermes still comes back empty, confirm the binary actually exists before troubleshooting PATH further:

ls ~/.local/bin/hermes

If that file is missing too, the install itself didn’t complete — re-run the installer rather than continuing to fight PATH.

Issue 2: Playwright/Chromium Install Fails During Setup

Symptom: Near the end of the installer (or when hermes doctor runs later), you see:

→ Installing browser engine (Playwright Chromium)...
⚠ Playwright browser installation failed — browser tools will not work.
⚠ Try running manually: cd ~/.hermes/hermes-agent && npx playwright install --with-deps chromium

Cause: Chromium’s install process needs more memory headroom than a 1 GB instance has free, and free-tier VPS images typically ship with zero swap configured. Everything else — the core agent, CLI, dependencies, config, and skills — installs and runs fine regardless; only the optional browser-automation tool is affected.

Fix: Add 2 GB of swap, then retry:

sudo fallocate -l 2G /swapfile
sudo chmod 600 /swapfile
sudo mkswap /swapfile
sudo swapon /swapfile
echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab
free -h   # confirm ~2.0Gi now shows on the Swap line

cd ~/.hermes/hermes-agent
npx playwright install --with-deps chromium

If you don’t need browser automation (scraping JS-heavy pages, visual checks), it’s fine to just leave this disabled and move on — it’s one tool out of Hermes’s full pool, not a dependency for the rest of the agent.

Issue 3: Wrong Gateway Subcommand Order

Symptom: Following instructions that say hermes setup gateway, you get a “command not found” or usage error, even though hermes itself works fine.

Cause: The documented subcommand order is hermes gateway setup, not hermes setup gateway. It’s an easy mistake to copy from a source that got the order backwards — including, in this case, another AI assistant’s generated instructions.

Fix: Use the order from the official README:

hermes gateway setup      # configure Telegram, Discord, Slack, WhatsApp, or Signal
hermes gateway start      # run the messaging gateway
hermes gateway restart    # restart it after a config change (e.g. new allowlist entry)

If you’ve seen references to hermes gateway install, hermes gateway status, or a built-in systemd --user service manager with loginctl enable-linger — none of that appears in the official GitHub README or the docs site as of this writing. Don’t build automation around subcommands you haven’t confirmed exist; check hermes gateway --help or the docs directly first.

A General Rule for This Kind of Setup

Every one of these gotchas is small on its own, but they share a root cause worth naming: copy-pasted instructions — from a blog post, a chat assistant, or a stale doc — can be confidently wrong about exact command names, flags, or mechanisms that were never actually shipped. Before running something on a production box (or even a throwaway free-tier VPS), a quick --help or a check against the project’s own README catches this class of error before it costs you a debugging session.

For the full setup this troubleshooting guide builds on, see How to Deploy Hermes Agent on Oracle Cloud’s Free Tier.

Frequently Asked Questions

Why does 'hermes' say command not found right after installing it?

The installer adds ~/.local/bin to PATH by editing ~/.bashrc, but a shell session that was already open won't pick that up automatically. Run `source ~/.bashrc` (or open a fresh SSH session) and confirm with `which hermes`.

Why does the Playwright/Chromium install fail during Hermes setup?

On a 1 GB RAM instance (like Oracle Cloud's free VM.Standard.E2.1.Micro shape) there usually isn't enough memory headroom to install Chromium, and no swap is configured by default. Add a 2GB swap file first, then retry `npx playwright install --with-deps chromium`.

Is it 'hermes gateway setup' or 'hermes setup gateway'?

It's `hermes gateway setup`, followed by `hermes gateway start`. Those are the only two gateway subcommands documented in the official README — running them in the reversed order, or expecting extra subcommands like `gateway install` or `gateway status`, will fail because those aren't part of the documented CLI.

Can I trust AI-generated Hermes Agent commands from ChatGPT or other assistants?

Treat them as a draft, not a fact. Cross-check any specific subcommand, flag, or claimed behavior against the official GitHub README or docs site before running it — assistants (this one included) can confidently suggest commands or mechanisms that were never actually shipped.

#Hermes Agent #Troubleshooting #Nous Research #VPS #Oracle Cloud
Share:
AI Integration & GPU Platforms

Need help with AI Integration & GPU Platforms?

Need help deploying AI/ML platforms? Get expert consulting on OpenShift AI, GPU orchestration, and MLOps.

Learn more about AI Integration & GPU Platforms

Want to operate this yourself, in production?

Take the free AI Platform Engineer Readiness Scorecard to see which skills transfer — then build a production-shaped AI platform in the 4-week Bootcamp.

Take the Scorecard →
Luca Berton — AI & Cloud Advisor, Docker Captain

Luca Berton

AI & Cloud Advisor · Docker Captain · KubeCon Speaker

15+ years in enterprise infrastructure. Author of 8 technical books, creator of Ansible Pilot (1M+ YouTube views, 648K site users). Former Red Hat engineer. Speaker at KubeCon EU 2026 and Red Hat Summit 2026.

Free 30-min AI & Cloud consultation

Book Now