> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fliiq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> From zero to running in 5 minutes

## Prerequisites

* Python 3.12+
* An API key from [Anthropic](https://console.anthropic.com/), [OpenAI](https://platform.openai.com/), or [Google Gemini](https://aistudio.google.com/) — or a [self-hosted LLM](/guides/configuration#self-hosted-llms)

## Install

```bash theme={null}
pip install fliiq
```

## Initialize

```bash theme={null}
fliiq init
```

This creates `~/.fliiq/` with a `.env` template. Open it and paste your API key:

```bash .env theme={null}
ANTHROPIC_API_KEY=sk-ant-...
```

At least one LLM API key is required. If multiple are set, Fliiq uses Anthropic first, then OpenAI, then Gemini. Use `--provider` to choose a specific provider:

```bash theme={null}
fliiq --provider openai
```

See [Configuration](/guides/configuration#provider-switching) for persistent defaults and self-hosted LLM setup.

## Verify Setup

```bash theme={null}
fliiq doctor
```

`doctor` checks your API keys, SOUL.md, playbooks, MCP servers, and skill directories. Fix any issues it reports before continuing.

## Run Something

```bash theme={null}
# Single-shot command
fliiq run "what time is it in Tokyo"

# Build something
fliiq run "create a Python script that generates the first 20 fibonacci numbers"
```

Fliiq spins up an agent loop, selects the right tools (`get_current_time`, `write_file`, `shell`), executes them, and returns results.

## Start an Interactive Session

```bash theme={null}
fliiq
```

This launches the REPL where you can have multi-turn conversations. The agent remembers context within the session. For a full-screen TUI, use `fliiq tui`.

### Commands

| Command   | Action                                      |
| --------- | ------------------------------------------- |
| `/mode`   | Cycle mode (plan → supervised → autonomous) |
| `/status` | Show session info                           |
| `/clear`  | Reset conversation history                  |
| `/help`   | Show available commands                     |
| `/exit`   | Exit chat                                   |

## Execution Modes

Fliiq has three execution modes that control how much autonomy the agent has:

| Mode           | Behavior                                                            |
| -------------- | ------------------------------------------------------------------- |
| **Plan**       | Agent creates a plan, waits for your approval before executing      |
| **Supervised** | Agent executes step-by-step, pausing at each tool call for approval |
| **Autonomous** | Agent executes fully without interruption                           |

```bash theme={null}
fliiq run "refactor the auth module" --mode plan
fliiq --mode supervised
```

## What Just Happened?

When you ran a command, Fliiq:

1. Loaded your API key and initialized the LLM provider
2. Assembled a system prompt (SOUL.md + mode instructions + available skills)
3. Entered the agent loop — the model picks tools, executes them, evaluates results, and continues until done
4. Saved an [audit trail](/guides/audit) to `~/.fliiq/audit/`

## Next Steps

<CardGroup cols={2}>
  <Card title="Configuration" icon="gear" href="/guides/configuration">
    Set up Gmail, Twilio, Telegram, and other integrations.
  </Card>

  <Card title="Execution Modes" icon="sliders" href="/guides/modes">
    Understand plan, supervised, and autonomous modes in depth.
  </Card>

  <Card title="Skills" icon="puzzle-piece" href="/skills/overview">
    Browse the 57 built-in skills and learn how to create your own.
  </Card>

  <Card title="Daemon & Jobs" icon="clock" href="/daemon/overview">
    Set up scheduled automations that run in the background.
  </Card>
</CardGroup>