Skip to content

Quick Start

Get your agent running in three steps.

Step 1: Install

git clone https://github.com/langoai/lango.git
cd lango
make build

See Installation for detailed instructions, including optional C compiler setup required for CGO.

Step 2: Onboard

Run the guided setup wizard:

./bin/lango onboard

Or start from a preset template to skip manual feature selection:

./bin/lango onboard --preset <name>
Preset Description
minimal Basic AI agent (quick start)
researcher Knowledge, RAG, Graph (research/analysis)
collaborator P2P team, payment, workspace (collaboration)
full All features enabled (power user)

See Configuration Presets for full details on what each preset enables.

The wizard walks you through five steps:

1. Provider Setup

Choose an AI provider and enter your API credentials.

Provider What You Need
OpenAI API key
Anthropic API key
Gemini API key
Ollama Local Ollama server URL

2. Agent Config

Configure your agent's model settings:

  • Model -- Select which model to use (e.g., gpt-4o, claude-sonnet-4-20250514)
  • Max Tokens -- Maximum response length
  • Temperature -- Creativity level (0.0 = deterministic, 1.0 = creative)

3. Channel Setup

Connect at least one messaging channel:

  • Telegram -- Bot token from @BotFather
  • Discord -- Bot token from the Discord Developer Portal
  • Slack -- Bot token and App token from the Slack API dashboard

4. Security & Auth

Configure security features:

  • Privacy Interceptor -- Filter sensitive content from AI requests
  • PII Protection -- Automatically detect and redact personally identifiable information

5. Test Config

The wizard validates your configuration by checking:

  • Provider connectivity
  • Channel authentication
  • Configuration consistency

Tip

You can re-run lango onboard at any time to update your configuration. It loads your existing settings as defaults.

Step 3: Serve

Start the agent server:

./bin/lango serve

Your agent is now running and connected to your configured channels. Send a message through Telegram, Discord, or Slack to start a conversation.

Interactive TUI

Run lango without arguments to start the standalone mission workbench. If the active profile is incomplete, that first screen points you to lango onboard, lango settings, and lango doctor. Once the profile is ready, the same surface switches to starter prompts and the Enter / 1-3 quick-start path. Use lango cockpit for the full multi-panel operator dashboard:

lango

Health Check

Verify the server is running:

lango health

Or with curl:

curl http://localhost:18789/health

Beyond the Wizard

Full Configuration Editor

For access to all configuration options beyond what the onboard wizard covers:

lango settings

This opens an interactive TUI editor with every available setting.

Validate Configuration

Check your configuration for errors without starting the server:

lango config validate

Next Steps