Need help understanding the Walkable Climb and Height variables

So I’m using a recast graph and I was experimenting with scaling my agents (the entire gameObject) and I noticed that when I made agents bigger, they could walk over stuff that they couldn’t before.

I take it that this is controlled by the Walkable Climb and Height variables on the Graph, but how does it work exactly?

Are those variables meant to represent what percentage of an agent’s transform they can walk under and climb?

So my Walkable Climb is at 0.5, does that mean my agent can climb over anything less than half it’s transform’s size?

My biggest question is:

Is there a way to make larger agents behave the same way as smaller ones on a graph?

Hi

The walkable climb setting is in world units. It’s not affected by the runtime agent at all.
All settings in the recast graph view are used for generating the graph you see in the scene view. On that graph, every point on it should be a point where it is valid for the center of the character to be.
If you have a too big character and a small character using the same graph, this might lead to the large agent unintentionally climbing on top of obstacles when it is trying to follow its path.

You might be interested in this tutorial: Multiple agent types - A* Pathfinding Project