DAGs in Pylon: Classic Workflows

What you’ll accomplish

If you know Builder, you already know DAGs - this page maps the classic experience onto Pylon so you can keep building node workflows without missing a step.

What a DAG is

A DAG (directed acyclic graph) is the classic node workflow: you place nodes on a visual canvas, wire them together, and data flows through exactly the path you drew. Same canvas, same nodes, same behavior as Builder - Pylon renames the object and gives it a dedicated home.

Choose a DAG for fixed, repeatable pipelines; choose an Agent when the work is open-ended and the AI should decide the steps.

Creating a DAG

1. In the left navigation under Build, select DAGs, then Create DAG.
2. Give it a Name (the characters / # ? % $ aren’t allowed) and select Create DAG. You’ll land on the visual canvas.
3. Build as you always have - nodes, connections, test, publish. The node-by-node reference lives in the Builder documentation.

Prefer a head start? Build → Templates offers prepared workflows: “Choosing a template creates a new DAG and loads it onto the canvas.”

Where DAG things moved

You’re looking for	In Pylon
Your workflows	Build → DAGs (archived ones under the Archived toggle)
Prepared workflows	Build → Templates
Run monitoring	Monitor → Dashboard and Monitor → Activity
Usage statistics	Insights → Analytics (the page is titled Statistics; DAGs have their own tab)
Exposing DAGs as tools	Capabilities → MCP Servers - bundle DAGs as tool-calls any MCP client can call
Website chat embed	Deploy → Integrations → Websites, plus the Website widget tab when publishing a DAG

Plan limits

DAGs have their own plan quota, separate from Agents. At the cap, Create DAG tells you: “You have reached your DAGs plan limit. Upgrade your subscription to create more.”

Next steps

• Pylon Overview - the whole workspace
• Create your first Agent - the autonomous side of Pylon
• Builder documentation - canvas and node reference