← Back to home

Setup Guide

Everything you need to get RealizeOS running — from preparation to your first AI-powered conversation. Choose your edition below and follow the steps.

In This Guide

  1. Before You Start — What to Prepare
  2. Lite Edition Setup (Obsidian + Claude Code)
  3. Full Edition Setup (Docker Self-Hosted)
  4. Preparing for Your Setup Call ($499 tier)
  5. First Steps After Setup
  6. Troubleshooting

1. Before You Start — What to Prepare

Before purchasing or installing, gather these materials. Having them ready means you can go from purchase to working system in under an hour.

Preparation Checklist

Getting Your API Keys

Anthropic API Key (Claude models)

  1. Go to console.anthropic.com
  2. Create an account or sign in
  3. Navigate to API Keys and create a new key
  4. Add $5-10 credit (pay-as-you-go, no subscription required)
  5. Save the key — you'll need it during setup

Typical cost: $2-8/month depending on usage. Claude Sonnet handles most tasks.

Google AI API Key (Gemini Flash — optional but recommended)

  1. Go to aistudio.google.com/apikey
  2. Sign in with your Google account
  3. Click "Create API Key"
  4. Copy the key — done

Gemini Flash has a generous free tier. It handles quick tasks so Claude can focus on complex work.

2. Lite Edition Setup

~15 minutes

The Lite edition runs entirely through Obsidian and Claude Code. No servers, no Docker, no coding. Your knowledge base lives as markdown files that Claude reads and works from.

Prerequisites

Step-by-Step

  1. Download and unzip the RealizeOS Lite package you received after purchase.
  2. Open as Obsidian vault: Launch Obsidian → "Open folder as vault" → select the realize_lite/ folder.
    Important: Save in your Documents folder, not in a cloud-synced folder (OneDrive, iCloud).
  3. Open the setup guide: In Obsidian, open setup-guide.md. It walks you through:
    • Filling in your brand identity
    • Defining your voice rules
    • Customizing your agent team
    • Setting up your first skill workflow
  4. Launch Claude Code in the vault directory. Claude reads your CLAUDE.md file and becomes your AI team — aware of your brand, agents, and knowledge.
  5. Test it: Try asking Claude to: 
    Write a LinkedIn post about [your topic]
    Verify the Writer agent responds in your brand voice, then the Reviewer checks quality.

What you should see

3. Full Edition Setup

~30 minutes

The Full edition gives you the complete engine: multi-LLM routing, API server, tool integrations, and self-evolution. Run it locally or deploy to any server.

Prerequisites

Step-by-Step

Option A: Docker (recommended)

# 1. Clone the repository
git clone https://github.com/SufZen/realize-os.git
cd realize-os

# 2. Configure your environment
cp .env.example .env
# Edit .env — add your ANTHROPIC_API_KEY and GOOGLE_AI_API_KEY

# 3. Initialize from a template
python cli.py init --template consulting
# Other templates: agency, multi-venture, saas, ecommerce

# 4. Start everything
docker compose up

# Server runs at http://localhost:8080

Option B: Local Python

# 1. Clone and install
git clone https://github.com/SufZen/realize-os.git
cd realize-os
pip install -r requirements.txt

# 2. Configure
cp .env.example .env
# Edit .env with your API keys

# 3. Initialize
python cli.py init --template consulting

# 4. Start the server
python cli.py serve

Verify It Works

# Send a test message
curl -X POST http://localhost:8080/api/chat \
  -H "Content-Type: application/json" \
  -d '{
    "message": "Help me plan a content strategy for Q2",
    "system_key": "consulting",
    "user_id": "test-user"
  }'

Expected response

You should get a JSON response with an AI-generated strategy, routed through the appropriate agent (likely the Analyst or Strategist), written in the brand voice defined in your system's Foundations.

Optional: Connect Telegram

  1. Create a bot via @BotFather on Telegram
  2. Add the bot token to your .env file
  3. Set AUTHORIZED_USERS to your Telegram user ID
  4. Restart the server — your bot is live

4. Preparing for Your Setup Call

If you purchased the Setup Assistance package ($499), here's how to make the most of your 1-hour call.

Before the Call — Prepare These

What Happens During the Call

  1. Discovery (10 min) — We review your business, ventures, and goals
  2. Brand Configuration (15 min) — We set up your identity, voice rules, and brand standards
  3. Agent Setup (15 min) — We define and customize your AI team members
  4. First Workflow (15 min) — We build and test your first skill workflow end-to-end
  5. Testing & Q&A (5 min) — We verify everything works and answer questions

After the Call

5. First Steps After Setup

Your system is running. Now what? Here's a roadmap for your first week.

Day 1 — Foundations

Day 2-3 — Knowledge

Day 4-5 — Workflows

Week 2+ — Expansion

6. Troubleshooting

"Claude doesn't know my brand voice"

Make sure your F-foundations/brand-voice.md file has clear, specific rules. Vague instructions like "be professional" don't help. Instead: "Use short paragraphs (2-3 sentences). Start with data, not opinions. No exclamation marks."

"Docker won't start"

Check that Docker Desktop is running. Try docker compose down then docker compose up --build. Check the logs with docker compose logs -f.

"API key errors"

Verify your .env file has no spaces around the = sign. Check that your Anthropic account has credit. Test the key directly: curl https://api.anthropic.com/v1/messages ...

"Agents respond generically"

The knowledge base might not be indexed. Run the index command (/index in Telegram or python cli.py index). Also check that your agent definitions in A-agents/ have specific, detailed role descriptions.

Still stuck?

Join the community: Telegram Builders Group or email support@realizeos.ai

Ready to get started?

Choose Your Edition