Workflow Connections

When a workflow does not behave as expected, the problem is often how nodes are connected, not the nodes themselves.

This page helps you understand:

How workflow connections work
The most common connection mistakes
How to fix data flow and execution order issues

No programming knowledge is required.

How Workflow Connections Work

A workflow is a sequence of nodes connected together.

Each connection:

Sends data from one node to the next
Defines the order in which nodes run
Makes sure the right information arrives at the right step

If a connection is missing or incorrect, the workflow may:

Stop early
Run in the wrong order
Produce empty or incorrect results

First Things to Check

Before looking deeper, always verify:

All nodes are visually connected
No required input is left unconnected
Connections go from left to right (earlier → later)
There are no loops unless intentionally designed

If something looks disconnected, it usually is.

Nodes Are Not Passing Data

Nodes Look Connected but Nothing Happens

Typical symptoms:

A node runs but outputs nothing
The next node runs with empty data
Results are missing or incomplete

Most common reasons:

The previous node did not produce any data
The wrong output was connected
The connection goes to the wrong input

What to do:

Click the previous node and confirm it produces output
Reconnect the line to the correct input
Remove unused connections

Execution Order Problems

Nodes always run in the order defined by connections.

If a node runs too early, it may not receive the data it needs.

Correct Execution Flow

flowchart TD
    A[Start] --> B[Extract Data]
    B --> C[Process Data]
    C --> D[Use Result]
    class A awf-trigger
    class B awf-io
    class C awf-data
    class D awf-ok

Common Wrong Pattern

flowchart TD
    A[Start] --> C[Process Data]
    C --> B[Extract Data]
    class A awf-trigger
    class B,C awf-err

Fix:

Ensure every node that needs data is connected after the node that produces it
Avoid branching too early unless necessary

Missing or Empty Data

If a node receives empty data:

The extraction step may have failed
The page may not have loaded yet
The selector or target may not exist

Recommended fix: Add a Wait for Element node before extracting data.

Helpful links:

Too Many Connections

Large workflows with many connections are harder to maintain.

Signs the workflow is too complex:

Many lines crossing each other
Multiple branches feeding the same node
Difficult to tell where data comes from

Better approach:

Split the workflow into smaller steps
Keep each workflow focused on one task
Reuse simpler workflows when possible

See:

Performance Optimization

Visual Guide to Connections

The workflow editor with the Hash Text example: a sticky note explaining it, then five connected steps (Run now, Your text, SHA-256, SHA-512, Show the hashes). The toolbar with Editor, Logs and Executions tabs is at the top, the Test Workflow button at the top right, and zoom controls at the bottom left. — Add a step (node)
Editor, Logs and Executions tabs
Test Workflow: run it now
A step (node); drag from its handle to connect

Best Practices for Reliable Connections

Connect nodes only when data is actually needed
Keep workflows linear when possible
Avoid unnecessary branches
Test each step before adding the next one
Prefer clarity over compactness

A simple workflow is usually a reliable workflow.

When Connection Issues Persist

If the workflow still fails:

Test it on a simpler website
Temporarily disable other browser extensions
Remove nodes one by one to find the breaking point
Rebuild the workflow from scratch if needed

Summary

Most workflow connection issues come from:

Missing or incorrect connections
Nodes running in the wrong order
Empty data passed between steps
Overly complex workflows

In most cases, the fix is visual: check the lines, the order, and the data flow.

If everything connects clearly, the workflow usually works.

Ask Aria