Getting started

ARTA runs entirely on your machine: a Docker Compose stack for the platform and a local LLM through Ollama by default. No API keys, no cloud account, no data leaving your network.

Prerequisites

• Docker with the Compose plugin (docker compose version should work)

• Ollama running locally, with a model pulled:

  ollama pull qwen2.5:32b

  Or skip Ollama and use a cloud LLM key (Anthropic, OpenAI, Google Gemini) — see Configuration.

Quickstart

git clone https://github.com/AmeyaAI/OPEN-ARTA.git
cd OPEN-ARTA
cp .env.example .env          # defaults are Ollama-local; add keys only if you want cloud LLMs
docker compose up -d

Then open:

• UI — http://localhost:38088
• API — http://localhost:38087

First login

Pick one of two paths before the first start:

1. Real admin account — set both variables in .env; ARTA creates this admin when the users table is empty, and you should unset them after first login:

   ARTA_BOOTSTRAP_ADMIN_EMAIL=admin@example.com
   ARTA_BOOTSTRAP_ADMIN_PASSWORD=change-me

2. No-database demo — set ARTA_DEMO_MODE=1 and log in as admin@arta.dev / demo1234. Never use demo mode in production.

Next steps

• Your first test run — from a project to a truthful report. The fastest first win is the API-test path from an OpenAPI spec.
• Installation — everything the quickstart glossed over: services, ports, environment, telemetry.