Quick Start

This guide gets you running with a local Komand instance using Docker Compose.

Prerequisites

• Docker and Docker Compose v2+
• 4 GB RAM minimum

1. Clone the Repository

git clone https://github.com/komand-ai/komand-server.git
cd komand-server

2. Start the Stack

docker compose -f docker/docker-compose.yml up -d

This starts four services:

Service	Port	Purpose
komand-postgres	5432	PostgreSQL 17 database
komand-seq	5341	Seq log viewer UI
komand-silo	11111, 30000	Orleans silo (grain host)
komand-gateway	5000	API gateway + web dashboard

3. Verify It’s Running

curl http://localhost:5000/health

You should get:

{ "success": true, "data": { "status": "healthy" } }

4. Send Your First Message

curl -X POST http://localhost:5000/api/webchat/message \
  -H "Content-Type: application/json" \
  -d '{
    "senderId": "user-1",
    "text": "Hello, Komand!",
    "displayName": "Alice"
  }'

This sends a message through the WebChat channel adapter. The default agent will process it and return a response.

5. Open the Dashboard

Navigate to http://localhost:5000 in your browser to access the web dashboard.

6. View Logs

Open the Seq log viewer at http://localhost:5341 to see structured logs from all services.

Development Mode

For local development without PostgreSQL, set the environment to Development:

DOTNET_ENVIRONMENT=Development

In development mode, Komand uses:

• In-memory grain storage (no database required)
• Localhost clustering (single-node)
• In-memory reminder service
• Swagger UI at /swagger

Next Steps

• Key Concepts — understand agents, sessions, skills, and channels
• Agents Guide — configure and manage AI agents
• REST API Reference — full API documentation
• Self-Hosted Deployment — production deployment guide