Execution Graph

Execution Graph#

Overview#

The Execution Graph is a powerful visual node-based automation system that allows you to create complex workflows without writing any code. Think of it as a visual programming language where you connect different nodes together to build sophisticated automations for your entire workflow.

With the Execution Graph, you can:

Automate repetitive tasks across your entire workflow
Create complex workflows using simple drag-and-drop operations
Build conditional logic and data processing chains
Integrate with external tools and services
Trigger automations based on pipeline events

Getting Started#

Check out the following guide to learn the basics of using the Execution Graph.

Getting Started

User Interface Guide#

Interface Overview#

The Execution Graph interface is designed for efficient workflow creation. Here's what each area does:

Central Node Graph

This is your main workspace where you'll create and connect nodes to build your automation workflows.

Output Log (Bottom Panel)

Shows real-time feedback and results from each node as your graph executes. This is invaluable for debugging and monitoring your automations.

Navigation Tab Bar (Top)

Navigate between nested node graphs and access the "Run" button to execute your current graph.

Right-Side Panel System

Three essential panels help you configure and manage your work:

Node Parameters Panel - Configure settings for the currently selected node
Graph Settings Panel - Set up graph-wide options like descriptions and callback triggers
Graph Library Panel - Save, load, and organize your node graph templates

Creating Nodes#

There are two convenient ways to add nodes to your graph:

Method 1: Tab Menu

Access the comprehensive node library through the tab menu system.

Method 2: Right-Click Context Menu

Right-click anywhere in the node graph for quick access to all nodes.

Connecting Nodes#

Creating connections between nodes defines the flow of your automation workflow. Here's how it works:

Basic Connection Process:

Click and drag from any output port (right side of a node)
Drop onto a compatible input port (left side of another node)
The connection line will appear, showing your data flow

Connection Rules:

Execution ports (flow control) can only connect to other execution ports
Cycles are prevented - you cannot create loops that would cause infinite execution
Most input ports accept only one connection, but some special nodes like the Merge node can accept multiple inputs

Advanced Connection Techniques:

Lazy Connection Mode

Hold ALT and drag with the right mouse button to enable "Lazy Connection" mode. This automatically connects the closest output port from your starting node to the closest input port of your target node. Perfect for quick connections!

Disconnecting Nodes#

Remove connections when you need to modify your workflow:

Method 1: Direct Deletion

Select any connection line and press Delete to remove it.

Method 2: Slicer Tool

Press Y to activate the slicer tool, then drag across connection lines to cut them quickly.

Method 3: Node Shaking

Rapidly move a node in different directions to automatically disconnect all its connections. This is useful for quickly isolating a node.

Removing Nodes#

Simple Deletion:

Select any node and press Delete to remove it from your graph.

Working with Nested Graphs#

Some nodes support nested node graphs, allowing you to create modular, reusable components:

Supported Node Types:

Group Node - Create reusable sub-workflows
For Each Node - Define iterative operations
Other container nodes - Various specialized containers

Accessing Nested Graphs:

Double-click any node that supports nesting to open its internal graph in a new tab. Use the tab bar to navigate between your main graph and nested sub-graphs.

Saving and Loading Graphs#

File Management:

Use the File Menu to save and load your node graphs.

Keyboard Shortcuts#

Master these shortcuts to work efficiently with the Execution Graph:

Node Management

Tab - Open the node creation menu
Delete - Remove selected nodes and connections
Y - Activate the connection slicer tool

Clipboard Operations

Ctrl + C - Copy selected nodes to clipboard
Ctrl + V - Paste nodes from clipboard

File Operations

Ctrl + N - Create a new blank graph
Ctrl + O - Open an existing graph file
Ctrl + S - Save the current graph
Ctrl + Shift + S - Save graph with a new name

Window Management

Ctrl + Shift + N - Open a new Editor window

Edit Operations

Ctrl + Z - Undo last action
Ctrl + Y - Redo last undone action
Ctrl + A - Select all nodes in the graph

Application Control

Alt + F4 - Close the editor
F5 - Execute the current graph

Node State Management

D - Toggle disabled/enabled state of selected nodes
I - Show detailed information for selected nodes

Layout and Organization

F - Frame selected nodes (or all nodes if nothing selected)
L - Auto-layout selected nodes (or all nodes if nothing selected)
Ctrl + L - Layout selected nodes horizontally
Ctrl + Alt + L - Layout horizontally with even spacing
Shift + L - Layout selected nodes vertically
Shift + Alt + L - Layout vertically with even spacing

Positioning Aids

S - Toggle snap-to-nodes when moving nodes
Shift + S - Toggle snap-to-grid when moving nodes

Advanced Connection Modes

Alt + Right Mouse - Lazy Connect Mode (automatic connection)
Ctrl + Alt + Right Mouse - Lazy Connect Mode (replace existing connections)
Ctrl + Shift + Alt + Right Mouse - Lazy Connect Mode (create multiple connections)

Running Your Automations#

There are several ways to execute your node graphs, each suited for different workflow needs:

In the Editor#

Full Graph Execution

Click the Run button in the toolbar to execute your entire graph. The execution will:

Start from the "Input" node
Follow the execution connections in sequence
Process each connected node according to your workflow logic
Display results in the Output Log

Selective Node Execution

For testing and debugging, you can execute specific parts of your graph:

Right-click any node in your graph
Select "Execute" from the context menu
The system will automatically execute that node and all its required input nodes

This is perfect for testing individual components or debugging specific workflow sections.

Automatic Execution with Callbacks#

Set up your graphs to run automatically when specific pipeline events occur:

Configuring Callbacks

In the Graph Settings panel, you can define callback triggers for automatic execution. Your graphs can respond to Prism Pipeline events such as:

When a new shot is created
When an asset gets exported
When files are published
When project states change
Other custom pipeline events

This powerful feature allows you to create responsive automations that trigger exactly when needed in your pipeline workflow.

Programmatic Execution#

Python API

Execute graphs programmatically using Python:

pcore.getPlugin("ExecutionGraph").executeGraphFromFile(graphPath)

This allows you to integrate Execution Graph automations into custom scripts, other plugins, or external applications.

Graph Libraries and Templates#

The Graph Library system helps you organize, share, and reuse your automation workflows:

Saving and Loading Graphs

Use the Graph Library panel to manage your workflow templates and share them across projects.

Prism Online Library

Expand your automation capabilities by accessing community-created graphs:

Click the "More" button in the Graph Library panel
Browse and download additional graphs from the Prism Online Library
Integrate community solutions into your workflows

Project-Specific Libraries

Store and organize graphs within your project libraries:

Save commonly used workflows for easy team access
Create project-specific automation templates
Important: Remove the .json file filter from your library in the Project Settings to enable saving node graphs to custom project libraries

Extending with Custom Nodes#

Create Your Own Node Types#

Method 1: Using the Interface
Go to File → Create Node Type
Follow the guided creation process
Define your node's inputs, outputs, and functionality

Method 2: JSON Definition
Copy the JSON node definition from an existing node
Modify it to create your custom functionality
Save it in a location where Prism scans for node definitions

Custom Node Loading System#

Prism uses a flexible system to load custom node definitions:

Default Location

Standard nodes are loaded from:

Resources\Nodedefs\standard_nodedefs.json

User-Specific Nodes

Personal custom nodes can be placed in:

%USERPROFILE%\Prism2\Presets\ExecutionGraph\NodeDefs\

All .json files in this directory will be automatically scanned.

Advanced Configuration

For complex setups, use environment variables:

PRISM_EG_NODEDEF_PATHS - Specify additional .json files to load
PRISM_EG_NODEDEF_SEARCH_PATHS - Define directories to scan for node definition files

All specified search paths will be recursively scanned for JSON files containing node definitions.

Benefits of Custom Nodes:

Extend Execution Graph functionality for specific pipeline needs
Create reusable components for complex operations
Share specialized nodes across your team or the community

Quick Node Creation Tip:

You can paste custom node type definitions directly into the node graph to instantly create nodes from JSON definitions stored in your clipboard.

Node Reference#

Node Reference

Configuration and Environment Variables#

Fine-tune the Execution Graph plugin behavior using these environment variables:

Environment Variable Configuration#
Variable Name	Description
PRISM_EG_GRAPH_FOLDERS	Specify additional folders that will appear in the Graph Library for loading node graphs. Useful for organization-wide template sharing.
PRISM_EG_NODEDEF_PATHS	Define specific file paths containing custom node definitions. Add multiple paths separated by your system's path separator.
PRISM_EG_NODEDEF_SEARCH_PATHS	Set directories that will be recursively searched for node definition files. Great for organizing custom nodes by category.
PRISM_EG_SHOW_SHADOWS	Control visual performance: `1` (default) enables attractive node shadows, `0` disables them for better performance in very large graphs.