Workflow Basics#

A Canopy workflow is a graph. Data moves through edges. Nodes hold the logic.

Use this guide to build a small manual workflow and see the core model without external APIs or AI providers.

What you will build#

Manual Trigger → Constant → Set Workflow Output

The workflow starts when you run it, emits a fixed value, and returns that value as the final output.

1. Create a workflow#

Open Workflows and create a new workflow.

Give it a short name, such as First Workflow. Start from an empty graph.

2. Add a Manual Trigger#

Add a Manual Trigger node.

Manual triggers are useful while building because nothing runs until you start the workflow yourself.

3. Add a Constant node#

Add a Constant node and connect the Manual Trigger to it.

Set the constant value to something simple:

{
  "status": "ready",
  "source": "workflow-basics"
}

The Constant node gives you predictable data to inspect while learning the editor.

4. Add a Set Workflow Output node#

Add a Set Workflow Output node and connect the Constant node to it.

Set Workflow Output stores the value returned by the run. It is not a terminator or success marker.

5. Run and inspect#

Save the workflow, then run it from the manual trigger.

Inspect each node's output. You should see the same object move from Constant into Set Workflow Output.

The important part#

Canopy workflows are explicit. Edges carry data between nodes, and nodes decide what happens to the data they receive.

When a workflow gets complicated, this stays true. Branches, loops, AI nodes, table writes, document creation, and integrations all follow the same model.

Next steps#

Try replacing the Constant node with:

• Template to render text from input data
• JavaScript Script for custom logic
• HTTP Request to call an external API
• Document Create to turn output into a document