# Flows

> How do I list and push flows using the Windmill CLI?

## Listing flows

The `wmill flow` list command is used to list all flows in the remote workspace.

```bash
wmill flow
```

## Pushing a flow

Pushing a flow to a Windmill instance is done using the `wmill flow push` command.

```bash
wmill flow push <file_path> <remote_path>
```

### Arguments

| Argument      | Description                                                    |
| ------------- | -------------------------------------------------------------- |
| `file_path`   | The path to the flow file to push.                             |
| `remote_path` | The remote path where the flow specification should be pushed. |

### Examples

1. Push the flow located at `path/to/local/flow.yaml` to the remote path `f/flows/test`.

```bash
wmill flow push path/to/local/flow.yaml f/flows/test
```

## Creating a new flow

The wmill flow bootstrap command is used to create a new flow locally.

```bash
wmill flow bootstrap [--summary <summary>] [--description <description>] <path>
```

### Arguments

| Argument   | Description                          |
| ---------- | ------------------------------------ |
| `path`     | The path of the flow to be created.  |

### Examples

1. Create a new flow `f/flows/flashy_flow`

```bash
wmill flow bootstrap f/flows/flashy_flow
```

## Running a flow

Running a flow by its path is done using the `wmill flow run` command. Logs are streamed step-by-step with labeled headers showing each module's ID and summary. For-loop iterations are tracked individually as they complete.

```bash
wmill flow run <remote_path> [options]
```

### Arguments

| Argument      | Description                     |
| ------------- | ------------------------------- |
| `remote_path` | The path of the flow to be run. |

### Options

| Option         | Parameters | Description                                                                   |
| -------------- | ---------- | ----------------------------------------------------------------------------- |
| `-d, --data`   | `data`     | Inputs specified as a JSON string or a file using @filename or stdin using @- . Resources and variables must be passed using "$res:..." or "$var:..." For example `wmill flow run u/henri/message_to_slack -d '{"slack":"$res:u/henri/henri_slack_perso","channel":"general","text":"hello dear team"}'` |
| `-s, --silent` |            | Do not output anything other then the final output. Useful for scripting.      |

![CLI arguments](../../assets/cli/cli_arguments.png "CLI arguments")

:::tip Inspecting flow runs after completion
Use `wmill job get <job_id>` to see a hierarchical step tree with status and durations, or `wmill job logs <job_id>` to see aggregated logs from all steps. See [Jobs](./job.md) for details.
:::

## Update flow inline scripts lockfile

Flows inline script [lockfiles](../6_imports/index.mdx) can be updated locally using the [`wmill generate-metadata`](./generate-metadata.md) command:

```bash
wmill generate-metadata
```

This command handles scripts, flows and apps in a single pass. To only update flow lockfiles, use:

```bash
wmill generate-metadata --skip-scripts --skip-apps
```

:::info Legacy command
Prior to the unified command, this was done with `wmill flow generate-locks`. This command is now deprecated but still works.
:::

## Flow specification

You can find the definition of the flow file structure [here](../../openflow/index.mdx).

## Remote path format

```js
<u|g|f>/<username|group|folder>/...
```