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

# CLI reference

> Initialize FlareDB, start and stop local instances, and use the SQL shell from the command line.

## Overview

| Command               | Description                                                                                   |
| --------------------- | --------------------------------------------------------------------------------------------- |
| `flare init`          | Create the local directories and download the FlareDB engine and worker components. Run once. |
| `flare up`            | Start a local FlareDB instance on port `8099`.                                                |
| `flare down`          | Stop the running instance.                                                                    |
| `flare sql`           | Launch the interactive SQL shell against the local warehouse.                                 |
| `flare logs`          | Stream logs for the most recent job.                                                          |
| `flare logs <JOB_ID>` | Stream logs for a specific job.                                                               |
| `flare --version`     | Print the CLI version.                                                                        |
| `flare --help`        | List all available commands.                                                                  |

## Job logs

A `JOB-ID` is generated automatically for each submitted job. Use it to view logs for a specific job, or stream the most recent logs when you do not provide an ID.

```bash theme={null}
flare logs                  # Stream logs for the most recent job
flare logs <JOB_ID>         # Stream logs for a specific job
```

The `JOB-ID` is logged during job submission via the runner SDK.

## Related guides

* [Quickstart](/quickstart) - install the CLI and run a complete local WordCount example.
* [SQL shell](/sql) - query and manage FlareDB tables from the interactive shell.
* [Run a Beam pipeline](/usage) - submit your own pipeline jobs to a running instance.