Course 3

Rigging & Animation with Agentic AI

Make the same models move — and discover where the agent capability boundary actually is.

Rigging is the hardest stage for agents. Learn to push them to their limit, recognize where they break, and take over with judgment.

Choose your track

Track A (designers/artists): focus on judgment — reading agent output critically and steering it toward your intent.

Beginner

First prompts, basic constraints, reading agent output.

Your first armature prompt

ABC ~7 min

Ask the agent to "add bones" and it drops a tidy chain floating beside the model, or buried at the world origin, with every bone roll arbitrary — a skeleton that looks like a rig and deforms like a disaster the moment you move it.

Intermediate

Multi-step prompts, references, recovering from common failures.

Where auto-weights betray you

ABC ~7 min

Auto-weights give the agent a green light it can't read — the model binds, the agent reports success, and then an arm bends and the shoulder collapses like a crushed can, because weight quality is only visible in a pose the agent never struck.

Advanced

Repeatable workflows, quality gates, edge cases the agent fumbles.

Controllers, IK chains, and pole targets

ABC ~8 min

Ask for IK and the agent wires a chain that bends the knee backwards, snaps to the wrong pole, or quietly creates a dependency loop that makes the whole rig twitch — IK is a system of relationships, and the agent sets each relationship without seeing the others.

Master

The capability boundary: where you stop trusting the agent and take over.

The capability boundary — actual animation

ABC ~8 min

Ask the agent to "animate a walk cycle" and it produces keyframes that are technically valid and completely dead — feet that slide, no weight, no anticipation, mechanical timing — because expressive animation is the one thing in this whole pipeline the agent cannot yet author, only scaffold.

← All three courses