Lilly Tech Systems
Learn Sub Agents
Master the art of spawning specialized AI agents to divide, conquer, and parallelize complex tasks. Understand parent-child agent relationships used in Claude Code, Gemini CLI, and modern agentic tools.
Your Learning Path
Follow these lessons in order, or jump to any topic that interests you.
1. Introduction
What are sub agents? Understand parent-child agent relationships, why they exist, and where they are used in real tools.
2. How They Work
Agent spawning lifecycle, context passing, tool access, result aggregation, and memory isolation patterns.
3. Types of Sub Agents
Explore research, coding, testing, review, and planning agents. Learn when to use each type.
4. Implementation
Build sub agents in Claude Code, Python (Claude Agent SDK), and OpenAI Assistants with working examples.
5. Best Practices
Prompt design, cost management, error handling, debugging, and common pitfalls to avoid.
What You'll Learn
By the end of this course, you'll be able to:
Understand Agent Hierarchies
Grasp how parent agents spawn, manage, and collect results from specialized child agents.
Implement Sub Agents
Build working sub agent systems using Claude Code, Python SDK, and OpenAI Assistants.
Choose the Right Agent Type
Select between research, coding, testing, and planning agents based on your task requirements.
Optimize Agent Workflows
Design efficient agent chains that minimize cost while maximizing quality and throughput.