Quick Start

Get from zero to talking to your AI agent in under 5 minutes.

---

Step 1: Launch DotAgents

Open DotAgents after installing. On first launch, you'll see the onboarding wizard that walks you through initial setup.

Step 2: Configure an AI Provider

DotAgents needs at least one AI provider to work. Go to Settings > Providers and add your API key for one of:

Provider	Models	Best For
OpenAI	GPT-4o, GPT-4o-mini, o1, o3	General purpose, highest quality
Groq	Llama 3, Mixtral, Whisper	Fast inference, great for voice
Google Gemini	Gemini 2.0 Flash, Gemini Pro	Multimodal, large context
Local Models	Ollama, LM Studio	Privacy-first, offline capable

Enter your API key and select a model. For local models, set the custom base URL (e.g., http://localhost:11434/v1 for Ollama).

Step 3: Talk to Your Agent

Voice Input (Primary)

1. Hold Ctrl (macOS/Linux) or Ctrl+/ (Windows) to start recording
2. Speak your request naturally
3. Release to stop recording — your speech is transcribed and the agent responds
4. The response is automatically inserted into your active application

Agent Mode (with Tools)

1. Hold Ctrl+Alt to start recording in agent mode
2. Speak your request — the agent will use MCP tools to complete it
3. Release and watch real-time progress as the agent executes tools
4. Results are displayed or inserted automatically

Text Input

• Press Ctrl+T (macOS/Linux) or Ctrl+Shift+T (Windows) for direct text input

Step 4: Explore the Interface

Sessions View

The main interface shows your conversation history. Each session is a threaded conversation with your agent, complete with tool execution history and results.

Settings

Access settings from the sidebar to configure:

• General — AI provider, TTS voice, STT language
• Providers — API keys and model selection
• Capabilities — Enable/disable MCP servers and individual tools
• Agents — Create and manage agent profiles
• Loops — Set up recurring automated tasks

Panel Mode

DotAgents can run as a compact floating panel on your desktop — always accessible without taking up screen space.

Step 5: Add Your First MCP Tool

MCP tools extend what your agent can do. Go to Settings > Capabilities and add a tool server:

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/Documents"]
    }
  }
}

Now your agent can read and write files on your behalf. See MCP Tools for more servers.

---

What's Next

• Your First Agent — Create a specialized agent profile
• Voice Interface — Master the voice controls
• Agent Skills — Teach your agent new capabilities
• MCP Tools — Connect more tools and services