Data
Items, mapping, expressions, item linking. What a node receives, what it passes on, and how you put a value from an earlier step into a setting.
Concepts explain what happens underneath the canvas. Read them when a workflow runs but the result surprises you: a step ran ten times instead of once, a field came out empty, a branch went the wrong way, or an AI step answered from the wrong text.
Each concept page opens with a 20-second summary, shows a small run you can inspect, compares the options when there are several ways to do the job, and ends with the symptoms you’ll see when something is off.
Data
Items, mapping, expressions, item linking. What a node receives, what it passes on, and how you put a value from an earlier step into a setting.
Flow
Branching, repeating, waiting, failures. Which path a run takes, how work repeats, how a run pauses, and what happens when a step fails.
AI
Models, agents and tools, knowledge (RAG). Where a model runs, when an agent is worth it, and how to answer from your own documents.
If you’re new, read these five in order. Together they take about 20 minutes.
| You notice… | Read |
|---|---|
| A step ran many times when you expected once | Items |
| A field is empty, or shows the whole list | Mapping with expressions |
| A value from an earlier step belongs to the wrong record | Item linking |
| Nothing ran after an IF, Filter or Switch | Splitting and branching |
| You need to repeat steps, page by page or batch by batch | Looping |
| One failing step stopped the whole run | Error handling |
| An agent called the wrong tool, or never stopped | AI agents |
| An answer ignores your documents | RAG |
The glossary defines the terms used across the docs, and links each one to the page that explains it.
For the exact settings and outputs of a node, use the node reference. To practise a concept on a complete workflow, pick one of the recipes.