Nodes

Understand nodes, the building blocks of Draft & Goal workflows: inputs and outputs, node types like LLM and Web Scraper, configuration, and variables.

What is a node?#

A node is a single step in your workflow that performs a specific action. Think of nodes as individual LEGO pieces that you connect together to build something powerful.

Each node:

• Takes inputs (data to process)
• Performs an action (scrape, transform, analyze, etc.)
• Produces outputs (results to pass to the next node)

Anatomy of a node#

Connection points#

Every node has connection points:

• Input points (left side): Where data enters the node
• Output points (right side): Where results leave the node for the next step

Node types by function#

Configuring nodes#

Click on any node to open its configuration panel on the node.

Required fields#

Fields marked with a red asterisk (*) are required. The node won’t run without them. For example, the URL node requires a target address:

URL *: https://example.com

Using variables#

Instead of hardcoding values, use variables to pass data from other nodes:

URL: {{url}}

The {{text}} syntax means “use the output value from the node named text”.

Working with inputs#

Static values#

Enter values directly in the configuration. On an LLM node, for example:

Temperature: 0.7
Max Tokens: 1000

Dynamic values from other nodes#

Reference outputs from previous nodes:

Content: {{WebScraper}}

Combining static and dynamic#

You can mix both in text fields — for example in an LLM prompt that reads a Short Text value:

Summarize this article about {{topic}}:

Understanding outputs#

Every node produces an output that can be used by subsequent nodes. After a run, click a node to inspect its result in the output panel.

Output structure#

Most nodes output a JSON object with specific fields. A Web Scraper response might look like this:

{
  "status": "success",
  "data": "The scraped content...",
  "metadata": {
    "url": "https://example.com",
    "timestamp": "2024-01-15T10:30:00Z"
  }
}

Node categories#

AI Nodes#

Leverage powerful AI models for text generation, analysis, and more.

Data Source Nodes#

Connect to external data sources.

Tools Nodes#

Transform and manipulate data.

Control Flow Nodes#

Control how your workflow executes.

Best practices#

Keep nodes focused#

Each node should do one thing well. If a node is doing too much, split it into multiple nodes.

One mega-node that scrapes, 
summarizes, AND saves

Name your nodes#

Give nodes descriptive names by clicking on their title:

• ❌ LLM_0
• ✅ Summarize Article (on an LLM node)

Test incrementally#

After adding each node, run your workflow to verify it works before adding more.

Troubleshooting#

Node shows an error#

1. Check that all required fields are filled
2. Verify variable references are correct
3. Ensure previous nodes completed successfully

Output is unexpected#

1. Click on the node to see its output preview (see Build & Run mode for canvas testing)
2. Check the logs for detailed execution info
3. Verify your variable paths are correct — see the Variables guide and tools like JSON Path Extractor when you need a nested field

Next steps#