Hermes Workspace V2 Setup Guide (15 Minutes)

Q: Does Hermes Workspace V2 work with local models?

Yes, via Ollama or Atomic Chat.

Hermes Workspace V2 setup takes about 15 minutes if you know what you're doing, and this is the step-by-step guide that gets you from zero to running in one sitting.

This isn't a "what is Hermes Workspace" guide — for that, read the original Hermes Workspace post or the Hermes Agent Workspace overview.

This is purely the V2 setup walkthrough — install, configure, and ship.

What You Need Before Starting Hermes Workspace V2 Setup

A few prerequisites before you touch the installer.

You need an existing Hermes setup — if you don't have one, follow How To Setup Hermes Agent first.

You need a machine with 16GB+ RAM for V2 to run smoothly.

You need Node.js 20 or newer if you're doing the manual install.

You need Docker installed if you're going the Docker route.

You need a model provider configured — Claude, OpenAI, Gemini, or a local model via Ollama.

Total prep time: 5 minutes if you have everything, 30 minutes if you need to install prerequisites.

The 15-Minute Hermes Workspace V2 Setup Path

I'm going to walk you through the fastest path that actually works.

It's broken into three phases — install, configure, ship.

If you follow this in order, you'll be running in 15 minutes.

If you skip steps you'll be debugging for hours.

Don't skip steps.

Phase 1 — Install (5 Minutes)

There are three install paths.

The Hermes-assisted install is the easiest.

The Docker install is the most isolated.

The manual install is the most flexible.

I recommend Hermes-assisted unless you have a specific reason for the others.

Path A — Hermes-Assisted Install (Recommended)

Open your existing Hermes terminal.

Type the following message into the chat: "Can you set up Hermes Workspace V2 for me?"

Hermes guides you through the install.

It clones the repo, installs dependencies, sets up the gateway, configures the connection, and tells you the URL to open.

Total time: 5 minutes (most of which is waiting).

This is the recommended path.

Path B — Docker Install (For Isolated Setups)

Open a terminal.

Pull and run the V2 image:

docker run -d -p 8080:8080 hermesworkspace/v2:latest

Wait for the container to start (about 30 seconds).

Open http://localhost:8080 in your browser.

Total time: 2-3 minutes.

This is the cleanest path if you want Workspace isolated from your host machine.

Path C — Manual Install (For Custom Setups)

Clone the repo:

git clone https://github.com/NousResearch/hermes-workspace
cd hermes-workspace
npm install
npm run dev

Wait for the dev server to start.

Open the URL it prints in your browser.

Total time: 5-10 minutes depending on your internet.

This is the path for power users who want to modify the source.

Phase 2 — Configure (5 Minutes)

You're now in the V2 UI.

Default profile is loaded.

Time to actually make it useful.

Step 1 — Connect Your Hermes Gateway

Click Settings in the sidebar.

Click "Gateway".

Confirm V2 found your existing Hermes setup.

If it didn't, paste the path to your Hermes directory.

Click "Test Connection".

You should see a green checkmark.

Time: 30 seconds.

Step 2 — Add Your Model Provider

Still in Settings, click "Models".

Add your provider — Claude, OpenAI, Gemini, or Ollama.

Paste your API key (or local endpoint for Ollama).

Click "Test".

You should see a green checkmark.

Time: 1 minute.

Step 3 — Create Your First Profile

Click "Profiles" in the sidebar.

Click "New Profile".

Name it something useful — I started with "Research".

Pick which skills to enable from the 2,000+ available.

For a research profile I enable: web search, scraping, summarisation, citation, and memory.

Save the profile.

Time: 2 minutes.

Step 4 — Import Your Memories

If you're migrating from V1, click "Memory" in the sidebar.

Click "Import".

Point V2 at your V1 memory directory.

V2 reads and converts everything automatically.

If you're starting fresh, skip this step.

Time: 1 minute.

Step 5 — Set Context Budget

In your profile settings, set a max token budget.

I use 50,000 tokens per profile by default.

This stops a runaway conversation from blowing your model bill.

Time: 30 seconds.

Phase 2 done.

Want my exact V2 profile templates? Inside the AI Profit Boardroom, I share my full V2 profile library, skill packs, and context budgets. Plus weekly coaching calls + 3,000+ members. $59/mo locked rate. Get access here

Phase 3 — Ship (5 Minutes)

You have a working V2 install with a configured profile.

Time to actually use it for something real.

Test 1 — Chat With Your Agent

Click "Chat" in the sidebar.

Send a message: "Summarise the latest AI agent news."

The agent should respond using the skills you enabled.

Check the inspector panel to confirm tool calls are working.

Time: 1 minute.

Test 2 — Add A Memory

Click "Memory" in the sidebar.

Click "Add".

Add something like: "My voice is direct, no fluff, sentence-per-line."

Save.

Go back to chat and ask the agent to write something in your voice.

Confirm it uses the memory.

Time: 1 minute.

Test 3 — Create A Task

Click "Tasks" in the sidebar.

Click "New Task".

Add a task in Backlog: "Research the top 5 AI agent frameworks of 2026."

Drag it to "To Do".

The agent should pick it up and start working in the background.

Time: 1 minute.

Test 4 — Mobile PWA Install

Open the V2 URL on your phone (you'll need to expose it via something like ngrok or run it on your local network).

In iOS Safari: Share > Add to Home Screen.

In Android Chrome: tap the install prompt.

Open the PWA from your home screen.

You're now chatting with your agent from your phone.

Time: 2 minutes.

Phase 3 done.

You have a working Hermes Workspace V2 setup.

Community Q&A On The V2 Setup

I did a live Q&A walking through the V2 setup with the community.

Worth watching if you hit any setup issues.

Setup Issues You'll Probably Hit

Three issues come up most often.

Issue 1 — Gateway Won't Connect

V2 needs the Hermes gateway running.

If it can't connect, make sure your underlying Hermes setup is up.

Run hermes status in a terminal to confirm.

If Hermes itself isn't running, start it before opening V2.

Issue 2 — Skills Won't Load

If skills show as "errored" after enabling, your skill format might be wrong.

V2 uses a stricter JSON skill schema than V1.

Run the skill migration tool: hermes workspace migrate-skills.

It auto-fixes most format issues.

Issue 3 — Memory Import Fails

If V2 can't read your V1 memory directory, check permissions.

V2 needs read access to the directory.

On macOS, you might need to grant Full Disk Access to the Hermes app.

System Settings > Privacy & Security > Full Disk Access > Hermes.

These three cover 90% of setup issues.

My Recommended First-Day Profile Setup

I'd recommend creating these four profiles on day one.

Profile 1 — Research

Skills: web search, scraping, summarisation, citation, memory.

Memory: empty (build it as you research).

Context budget: 50,000 tokens.

Profile 2 — Writing

Skills: writing, editing, voice match, citation, memory.

Memory: your voice/style guide, banned phrases, tone references.

Context budget: 30,000 tokens.

Profile 3 — Code Review

Skills: code analysis, security check, refactor suggestion, test gen, memory.

Memory: your codebase conventions, banned patterns, architecture notes.

Context budget: 100,000 tokens.

Profile 4 — Admin

Skills: email triage, calendar, notes, memory.

Memory: your contacts, your preferences, your recurring tasks.

Context budget: 20,000 tokens.

These four cover 80% of what most knowledge workers do.

Customise from here.

Want the exact profile JSON files? The AI Profit Boardroom has my full V2 profile library ready to import, plus the 2-hour Hermes course and weekly coaching. Twin guarantee — if I don't double your output in 90 days, full refund. 3,000+ members. Join here

Optional Phase — Multi-Agent Setup

Once you have a few profiles working, the next step is multi-agent orchestration.

V2's profile system makes this much easier than V1.

Step 1 — Create A Lead Profile

Make a profile called "Lead" with the orchestration skills enabled.

Step 2 — Create Worker Profiles

Make profiles for each worker — researcher, writer, fact-checker, editor.

Step 3 — Configure The Swarm

In Settings > Swarm, set up the topology — hierarchical, mesh, or hybrid.

Add the worker profiles to the swarm.

Step 4 — Run A Test Task

Open the Tasks panel.

Add a task: "Research, write, fact-check, and edit a 1,000-word piece on AI agent frameworks."

The lead profile coordinates the workers automatically.

Watch the virtual office view to see them working.

The Hermes Agent Swarm post covers this in more depth.

Setup Checklist

Save this for your next install.

[ ] Hermes installed and running
[ ] 16GB+ RAM machine
[ ] Node 20+ or Docker installed
[ ] Model provider configured
[ ] V2 installed via Hermes-assisted/Docker/manual
[ ] Gateway connection verified
[ ] Model provider added
[ ] First profile created
[ ] Memories imported (if migrating)
[ ] Context budget set
[ ] Chat tested
[ ] Memory tested
[ ] Task tested
[ ] Mobile PWA installed (optional)

If you check all 14 boxes, your V2 setup is production-ready.

Setup For Specific Use Cases

A few use-case-specific setup tweaks.

For SEO Work

Add the SEO skill pack to your Research profile.

See Hermes SEO for the full SEO workflow.

For Content Production

Add the content skill pack to your Writing profile.

Memory should include your style guide and brand voice.

For Development

Add the dev skill pack to your Code Review profile.

Point the file panel at your project directory.

For Personal Productivity

Add the admin skill pack to your Admin profile.

Connect calendar and email if your model supports them.

Where To Go After V2 Setup

Once V2 is running, the next things to explore.

Hermes Kanban for the task board workflow.

Hermes Agent Swarm for multi-agent setups.

Hermes Second Brain for knowledge management.

Hermes MCP Server for external integrations.

Each one builds on the V2 foundation you just installed.

FAQ — Hermes Workspace V2 Setup

How long does Hermes Workspace V2 setup take?

About 15 minutes if you follow this guide and have prerequisites ready.

What's the easiest way to install Hermes Workspace V2?

The Hermes-assisted install — just ask Hermes to set up V2 for you.

Do I need Docker to install Hermes Workspace V2?

No, but Docker is the cleanest option for isolated setups.

What's the minimum RAM for Hermes Workspace V2?

16GB recommended, 8GB will work but feel sluggish.

Can I migrate my V1 memories to V2?

Yes, V2 imports V1 memory directories automatically.

Does Hermes Workspace V2 work with local models?

Yes, via Ollama or Atomic Chat — configure in Settings > Models.

Can I run Hermes Workspace V2 on my phone?

Yes, install the URL as a progressive web app on iOS or Android.

About Julian

I'm Julian Goldie — AI entrepreneur, SEO expert, and founder of the AI Profit Boardroom (3,000+ members). I help business owners scale with AI agents, automation, and SEO.