Quickstart

Get Ploston running in 5 minutes.

Prerequisites

• Python 3.12 or higher
• Terminal access

Step 1: Install Ploston CLI

pip install ploston-cli

Step 2: Create a Project Directory

mkdir my-ploston-project
cd my-ploston-project

Create ael-config.yaml:

workflows:
  directory: "./workflows"

logging:
  level: INFO

Step 4: Create Your First Workflow

Create the workflows directory and workflows/hello.yaml:

mkdir workflows

name: hello
version: "1.0"
description: A simple hello world workflow

inputs:
  name:
    type: string
    default: "World"

steps:
  - id: greet
    code: |
      name = "{{ inputs.name }}"
      result = f"Hello, {name}!"

output: "{{ steps.greet.output }}"

Step 5: Validate the Workflow

ploston validate workflows/hello.yaml

Expected output:

✓ Workflow 'hello' is valid

Step 6: Run the Workflow

ploston run workflows/hello.yaml

Expected output:

Hello, World!

Run with custom input:

ploston run workflows/hello.yaml -i name=Alice

Expected output:

Hello, Alice!

Step 7: Start as MCP Server (Optional)

Start Ploston as an MCP server for AI agent integration:

# Start with stdio transport (for AI agent integration)
ploston serve

# Or start with HTTP transport (for REST access)
ploston serve --transport http --port 8080

The workflow is now available as an MCP tool named workflow:hello.

What's Next?

You've successfully:

• ✅ Created a Ploston project
• ✅ Written a workflow
• ✅ Validated and executed it
• ✅ Started Ploston as an MCP server

Continue Learning

• First Workflow Tutorial - Build a more complex workflow
• Workflow Authoring Guide - Deep dive into workflows
• CLI Reference - All available commands

Try More Examples

• Web Scraping Example
• Data Processing Example
• API Integration Example