Workflows Overview¶

Workflows are the core of m9m - automated processes that connect nodes to accomplish tasks.

What is a Workflow?¶

A workflow is a directed graph of nodes that:

Starts with a trigger (manual, webhook, schedule)
Processes data through connected nodes
Outputs results or performs actions

┌─────────┐     ┌─────────┐     ┌─────────┐     ┌─────────┐
│ Trigger │────▶│  Fetch  │────▶│Transform│────▶│  Send   │
│         │     │  Data   │     │  Data   │     │  Alert  │
└─────────┘     └─────────┘     └─────────┘     └─────────┘

Workflow Structure¶

{
  "name": "My Workflow",
  "nodes": [...],
  "connections": {...},
  "settings": {},
  "active": true
}

Field	Type	Required	Description
`name`	string	Yes	Workflow name
`nodes`	array	Yes	List of nodes
`connections`	object	Yes	Node connections
`settings`	object	No	Workflow settings
`active`	boolean	No	Enable triggers

Workflow Lifecycle¶

Create → Edit → Test → Activate → Execute → Monitor
                 ↑                              ↓
                 └──────────────────────────────┘
                         (iterate)

Key Concepts¶

Nodes¶

Building blocks that perform operations:

Triggers: Start workflows (Webhook, Cron)
Actions: Perform tasks (HTTP, Database, Email)
Transform: Modify data (Set, Filter, Code)

Connections¶

Define data flow between nodes:

{
  "Source Node": {
    "main": [[{"node": "Target Node", "type": "main", "index": 0}]]
  }
}

Data Items¶

Data flows as items - JSON objects with optional binary:

{
  "json": {"field": "value"},
  "binary": {}
}

Expressions¶

Dynamic values using input data:

{{ $json.fieldName }}

Workflow Types¶

Manual Workflows¶

Started via API or CLI:

m9m run my-workflow.json

Webhook Workflows¶

Triggered by HTTP requests:

POST http://localhost:8080/webhook/my-endpoint

Scheduled Workflows¶

Run on a schedule:

{
  "type": "n8n-nodes-base.cron",
  "parameters": {
    "cronExpression": "0 9 * * *"
  }
}

Best Practices¶

Clear naming - Use descriptive node and workflow names
Error handling - Add filters to check for errors
Modular design - Break complex logic into smaller workflows
Test thoroughly - Test with sample data before activating
Monitor executions - Watch for failures and optimize