CLI Overview

The ScoutOS CLI allows you to interact with and update Scout workflows via CLI commands.

Getting Started

The ScoutOS CLI is a powerful tool that helps you create, manage, and deploy AI workflows. Here’s a quick overview to get you started:

1. Installation: First, install the CLI using the platform-specific instructions below.
2. Authentication: Run scout link to connect the CLI to your Scout account.
3. Create a Project: Use scout init to create a new Scout project.
4. Basic Commands:

   $
   scout --help          # View all available commands
   >
   scout workflows list  # List all workflows
   >
   scout workflows run   # Run a workflow

For detailed instructions, follow the setup guide for your operating system below.

Setup

macOS

You can install the ScoutOS CLI using one of the following methods:

• Automatic Installation:

  $
  /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/scoutos/homebrew-scoutos-cli/main/install.sh)"

• Manual Installation:

  $
  brew tap scoutos/scoutos-cli
  >
  brew install scoutos-cli

Now you should be able to use scout from your system! Try running scout --help to see if it works.

Linux

You can install the ScoutOS CLI using one of the following methods:

• Automatic Installation:

  $
  /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/scoutos/homebrew-scoutos-cli/main/install.sh)"

• Manual Installation:

  $
  brew tap scoutos/scoutos-cli
  >
  brew install scoutos-cli

Now you should be able to use scout from your system! Try running scout --help to see if it works.

Windows

1. Download the Windows executable from the Latest Release.

2. Move and authorize the downloaded file:

   $
   # Move & rename executable
   >
   move scout-cli.exe C:\Windows\System32\scout.exe

Now you should be able to use scout from your system! Try running scout --help to see if it works.

Note: You may have to grant the CLI permissions to write, read, and delete on your system.

$
icacls C:\Windows\System32\scout.exe /grant Everyone:F

To remove the CLI:

$
del C:\Windows\System32\scout.exe

Connect to Your Scout Account

Authenticate your local environment with:

$
scout link

This will securely connect your CLI to your Scout account.

Initialize a Scout Project

To start a new Scout project in your local directory, run:

$
scout init

The CLI will walk you through the setup process.

Run a Workflow

To run a specific workflow, use the following command:

$
scout workflows run <workflow_folder>

Replace <workflow_folder> with the folder name of the workflow you want to run.

Deploy a Workflow

To deploy a workflow, use the following command:

$
scout workflows deploy <workflow_folder>

AI Workflow Project Structure

A Scout AI Workflows as Code project follows a structured hierarchy:

/workflows
├── workflow_1/
│   ├── workflow.yml
│   ├── inputs/
│   │   ├── default.json
│   ├── meta.ts
│
├── workflow_2/
│   ├── workflow.yml
│   ├── inputs/
│   │   ├── default.json
│   ├── meta.ts
│
└── scout.config.ts

Key Components:

• workflow.yml – Defines each AI workflow.
• inputs/default.json – Stores default input parameters for the workflow.
• meta.ts – Metadata and additional configuration for the workflow.
• scout.config.js – Configuration settings for your Scout project.

Local Development

Clone and setup:

$
git clone git@github.com:scoutos/scout-cli.git
>
cd scout-cli

Install prerequisites:

• Install Deno
• Setup your development IDE environment

Now you should be able to run the code locally:

$
deno task dev --help
>
deno task dev workflows deploy --config ./examples/starter/workflows/source_mapping.yml

Now you should be able to use scout from your system! Try running scout --help to see if it works.

If you want to remove the CLI, you can run:

$
sudo rm /usr/local/bin/scout

Note: You may have to grant the CLI permissions to write, read, and delete on your system.