Action

Execute Workflow

The Workflow node (Virtual Node) unlocks ultimate modularity. It executes a separate workflow as a modular function, safely handling recursion up to 3 levels deep. It seamlessly maps parent item data down to the child via `nlink-virtual.input` and awaits output data from `nlink-virtual.output`.

Execute Workflow

Core / Action

⚠️

What can you do with Execute Workflow?

Dynamic Injection & Inheritance

Sub-workflows automatically merge evaluated parent Node Properties with input payloads, and even inherit API Credentials globally without reconfiguration.

Strict Array Mapping

Maps arrays seamlessly. Parent workflow items are fed securely into the sub-workflow's `Virtual Input`, and `Virtual Output` responses are flawlessly re-mapped into the parent execution iteration loop.

Safe Guarded Recursion

The execution engine safely wraps sub-workflow calls, preventing infinite loop crashes by strictly enforcing a maximum recursion depth block (Max Depth: 3).

Detailed Usage & Configuration

The Workflow (Sub-Flow) node is the cornerstone of advanced orchestration. It allows you to build modular architectures instead of massive, unmaintainable "spaghetti" canvases.

1. Modular Reusability

If you have five different workflows that all eventually require the same 12 steps to process a customer order, do not duplicate those 12 steps five times! Instead, build one "Process Order" Workflow. Then, use the Workflow node across the other five canvases to trigger it. This centralizes your logic—update the "Process Order" flow once, and all five parent workflows benefit immediately.

2. The "Tool Execution Link" Mechanic

This node is fundamentally identical to the underlying architecture powering the AI Agent node's tool-calling logic. When configuring the Workflow node, you explicitly select a target Sub-Workflow from your workspace. The items (data payload) entering this node are securely injected into the target workflow via its Virtual Input node. The parent workflow pauses execution, waits for the Sub-Workflow to finish routing data into its Virtual Output node, and seamlessly consumes the result array to continue running the primary Sequence.

💡 Pro Tip - Sub-Workflow Entry Points: You do NOT need to add an Execute Workflow node at the beginning of your Sub-Workflow! The Sub-Workflow operates just like a normal workflow. It only needs a standard trigger (e.g., Webhook, Manual Trigger) or simply an action node (like Set) as its starting point. When triggered, the engine automatically finds the Sub-Workflow's trigger/first node and injects the parent data directly into it.