Best Practices

Tips for effectively using Embedded Mode

Follow these guidelines to get the most out of Embedded Mode.

Use for Scene-Specific Logic

Embedded mode shines when you need to reference specific scene objects:

• Level triggers and events

• Scene-specific cutscenes

• Environmental interactions

• Tutorial sequences

• One-time scene setups

Rule of thumb: If the logic only makes sense in one specific scene, Embedded Mode is a good choice.

Keep It Simple

Embedded graphs should be focused and simple. For complex logic:

• Use Sub-State Machines referencing Asset state machines

• Keep scene-specific parts embedded, share common logic as assets

• Break large graphs into smaller, manageable pieces

Hybrid Approach

Embedded Runner
├─ Trigger: Scene-specific event
└─ Sub-State Machine: [Asset] Shared AI Behavior

This pattern lets you use scene references for triggers while reusing complex logic.

Document with Sticky Notes

Since embedded graphs can't be easily shared or viewed outside the scene, add Sticky Notes explaining:

• The purpose of the state machine

• Scene dependencies and references

• Any special conditions or requirements

Use Self and Target References

When possible, prefer Self and Target references over direct scene references:

Reference Type	Use When
Self	Referencing the Runner's own GameObject
Target	Referencing a passed-in target
Scene Reference	Only when you need a specific, unchanging scene object

Consider Future Changes

Before using Embedded Mode, consider:

• Will this logic need to be reused?

• Might this become a prefab later?

• Will multiple objects need similar behavior?

If yes to any of these, start with Asset Mode instead.

Backup Important Graphs

Embedded graphs are stored in the scene file. To protect your work:

1. Regularly export complex embedded graphs as assets

2. Use version control for your scenes

3. Consider using Asset Mode for critical game logic