Getting Started with JARVIS

March 23, 2026 · View on GitHub

This guide walks you from a fresh checkout to your first AI conversation in under ten minutes.

Prerequisites:

  • Docker Desktop (or Docker Engine + Compose v2)
  • An OpenAI, DeepSeek, or Anthropic API key
  • Git

1. Clone and Configure

git clone https://github.com/your-org/JARVIS.git
cd JARVIS
bash scripts/init-env.sh

init-env.sh generates a .env file with random passwords and encryption keys. Open .env and fill in at least one API key:

# .env — fill in one of these
DEEPSEEK_API_KEY=sk-...
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...

2. Start the Stack

docker compose up -d

Wait until all containers are healthy:

docker compose ps
# Every row should show "healthy" or "running"

Open http://localhost in your browser.

3. Register and Log In

  1. Click Register on the login page.
  2. Enter an email and password (minimum 8 characters).
  3. You are automatically logged in after registration.

4. Configure Your LLM Provider

  1. Click your avatar (top-right) → Settings.
  2. Under AI Model Config, select your provider (DeepSeek / OpenAI / Anthropic / ZhipuAI / Ollama).
  3. Choose or enter a model name.
  4. In the API Keys section, paste your API key and click Save.

JARVIS encrypts all API keys at rest using Fernet symmetric encryption.

5. Send Your First Message

  1. Click New Conversation in the sidebar.
  2. Type a message and press Enter.
  3. The assistant streams its reply token by token.

Try a tool call: Ask "What time is it?" — JARVIS invokes the datetime tool and returns the current time.

Next Steps