Quickstart

You installed HiveMind OS — nice. Let's get you from zero to a working AI agent in about two minutes.

1. The Setup Wizard

Open HiveMind OS. On first launch you'll see a Setup Wizard that walks you through everything step by step:

Step	What you do
Welcome	Click Get Started
Providers	Add at least one AI model provider (see options below)
Connectors	Optionally connect email, calendar, or chat services — or skip for now
Web Search	Optionally configure web search — or skip
Models	Optionally download the BGE Small embedding model for local knowledge search
Tour	Quick overview of key features — then you're in!

You only need to complete the Providers step to start chatting. Everything else can be configured later in Settings.

2. Connect a Model Provider

In the Providers step, pick a provider card and fill in the details:

Option A: OpenAI-Compatible API Key (simplest)

1. Select OpenAI Compatible
2. Paste your API key and base URL. Done.

Option B: Anthropic

1. Select Anthropic
2. Paste your Anthropic API key. Done.

Option C: GitHub Copilot

1. Select GitHub Copilot
2. Your browser opens the GitHub OAuth flow — approve access
3. Token is stored securely in your OS keychain. Done.

Option D: Ollama (Local)

1. Select Ollama (Local)
2. Make sure Ollama is running on your machine (default: http://localhost:11434)
3. Pick which models to use. Done — fully offline.

TIP

You can add more providers later in Settings → Providers. HiveMind OS routes between them automatically based on task, cost, and data sensitivity.

3. Your First Conversation

Once connected, you land in the Chat view with a greeting from your agent. Let's try something practical.

Step 1 — Ask a real business question. Type (or copy-paste):

I need to write a professional email to a client explaining that their
project will be delayed by two weeks. Keep it empathetic and solution-focused.

Step 2 — Watch the agent work. You'll see HiveMind OS thinking in real time:

🧠 Planning: I'll draft a professional, empathetic email explaining the delay...

✍️ Drafting email with a clear explanation, apology, and revised timeline...

✅ Done — here's your email draft.

Step 3 — Get a polished email draft. The agent delivers a ready-to-send email — professional tone, empathetic language, and a clear next step for the client. Copy it, tweak it if you like, and send.

Now try something else:

Summarize the key points from this article: https://example.com/your-article

The agent fetches the page, reads the content, and gives you a clean summary with the key takeaways — no need to read the whole thing yourself.

4. What Just Happened?

Behind the scenes, HiveMind OS:

• Understood your request — figured out what you needed and the best way to help
• Chose the right approach — drafting, researching, analyzing, or a combination
• Used tools when needed — web search, file reading, email, and more
• Kept your data private — nothing left your computer unless you allowed it

That's the core of HiveMind OS: an AI assistant that thinks, acts, and respects your privacy — automatically.

5. What's Next?

You're up and running! Here's where to go from here:

For Business Users

• Use Cases — Ready-made automations for customer support, meeting prep, daily briefings, and more
• No-Code Guide — Do everything from the app — no terminal, no config files
• First Five Minutes — Quick tips to get the most out of HiveMind OS
• Privacy & Security — Learn how your private data stays private

For Developers

• Concepts → Agentic Loops — Understand how HiveMind OS reasons and plans
• Concepts → Knowledge Graph — See how memory works across sessions
• Concepts → Tools & MCP — Connect external tools like GitHub, Slack, or databases