---
title: Quick Start
url: https://arcana.otnelhq.com/docs/quickstart
---

# Quick Start



Get Arcana running in under a minute. This guide walks you through installation, provider setup, and your first session.




## Step 1 — Install



```bash
# One-shot (downloads binary on first run)
npx arcana-ai@latest

# Or install globally
npm install -g arcana-ai
arcana --version
```



Package name is `arcana-ai`; the binary is always `arcana`.




## Step 2 — Get Started



Arcana works in two modes:



### Bring Your Own Key (BYOK) — Free



Use your own provider API keys directly. Arcana detects them automatically:



```bash
# OpenAI
export OPENAI_API_KEY=sk-...

# Anthropic
export ANTHROPIC_API_KEY=sk-ant-...

# Google Gemini
export GEMINI_API_KEY=...

# Or any other supported provider
# See /docs/providers for the full list
```



### Arcana License — Proxy & Cloud Features



For the Arcana proxy (hosted API endpoint), cloud memory sync, and Gateway, you need an Arcana license. This is separate from provider API keys:



```bash
# Arcana license for proxy/cloud features
export ARCANA_LICENSE=your-license-key

# Proxy uses your license, not individual provider keys
arcana run --proxy "explain this codebase"
```



Your license covers proxy access. BYOK mode (direct provider keys) is always free.




## Step 3 — Check Health



```bash
arcana doctor
```



Runs 8 health checks — Bun runtime, node_modules, config file, API key, models cache, skills cache, bridge config, and Arcana home. Fix any issues it reports before proceeding. Add `--json` for machine-readable output or `--web` to also verify the enterprise web dashboard.




## Step 4 — Run



```bash
# Interactive TUI session
arcana

# One-shot prompt
arcana run "explain this codebase"

# Or just:
arcana "review open PRs for bugs"
```



Useful `run` flags:



| Flag | What it does |
| --- | --- |
| `-m, --model <model>` | Override the model for this run (e.g. `anthropic/claude-sonnet-4-6`) |
| `-p, --provider <name>` | Override the provider for this run |
| `-r, --resume <id>` | Resume a previous session by ID |
| `--skill <name>` | Activate a skill at session start |
| `--sandbox [dir]` | Isolate the agent to a root directory (add `--sandbox-net` for network) |
| `--proof` | Record a verifiable RunProof of the session |




## Step 5 — Explore



Inside the TUI, try these slash commands:



| Command | What it does |
| --- | --- |
| `/help` | List all available commands |
| `/skills` | Browse and activate skills |
| `/memory` | Search your memory database |
| `/compact` | Manually compact the session |
| `/providers` | Switch model providers mid-session |




## Optional Setup




### Console Login (Sync & Proxy)



```bash
arcana console login
```



Opens a browser for device-flow authentication at [arcana.otnelhq.com](https://arcana.otnelhq.com). Enables cloud memory sync and the proxy API.




### Trust Your Workspace



```bash
arcana trust
```



Required if your project has local plugins, tools, or MCP servers. Without trust, project-level extensions are blocked.




### Schedule Automation



```bash
# Run a code review every 4 hours
arcana cron add --name "review PRs" --schedule "0 */4 * * *" --prompt "review open PRs for bugs"

# Start the cron daemon
arcana cron start
```



See [Cron](/docs/cron) for full details.




### Chat Gateway



```bash
arcana gateway
```



Run Arcana as a bot on Telegram, Discord, Slack, or WhatsApp. See [Gateway](/docs/gateway) for setup instructions.




## What's Next?



- [Providers](/docs/providers) — Set up your preferred LLM provider
- [Model Recommendations](/docs/models) — Choose the best model for your task
- [Skills](/docs/skills) — Load domain expertise on demand
- [Memory](/docs/memory) — Persistent, searchable knowledge base
- [Configuration](/docs/configuration) — All config options and env vars
- [Examples](/docs/examples) — Real-world workflows and patterns