One agent is useful. A team of agents that coordinate with each other is transformative. But getting multiple agents to work together without duplicating effort, overstepping boundaries, or producing inconsistent results takes some thought.

​

What you’ll learn

• How to use skills to share knowledge and procedures across agents so they stay consistent
• When to break work into subagents vs. keeping everything in one agent
• How to set permissions and app policies so each agent only accesses what it needs
• Patterns for routing work between agents without losing context
• Common mistakes when scaling from one agent to many, and how to avoid them

​

Who this is for

Anyone who has built a few agents and is starting to think about how they work together. You should be comfortable creating agents and connecting tools. We’ll build on that foundation to cover multi-agent coordination.