Context Engineering for Coding Agents: Hands-on Lab + Agent Build-Off
Claude Code is powerful out of the box. But there's a gap between using it and mastering it.
This hands-on workshop teaches you to architect agent systems that scale. You'll move beyond basic usage to design custom subagents, optimize context windows, and build configurations that work in production.
Format: Individual diagnostic labs + team competition with live leaderboard
WHAT YOU'LL DO
Lab 1: Diagnose & Optimize (Individual)
Fix a flawed agent configuration. Optimize CLAUDE.md, add rules, improve skills, extend hooks. Scored challenges with measurable results.
Lab 2: Team Build-Off (Competition)
Design a subagent architecture to upgrade a REST API. Teams compete on a live leaderboard. Best architecture wins.
Plus: Prosus keynote on deploying agentic coding at scale.
What You'll Leave With
✅ How to improve rules, skills, hooks, and CLAUDE.md
✅ How to design useful subagents for real engineering tasks
✅ When to use one agent, subagents, or agent teams
✅ How to think about context engineering in a more practical way
WHO SHOULD ATTEND
Engineers and AI builders working with coding agents who want to move from basic usage to production-grade systems.
Prerequisites:
Command line comfort and ability to read/write Python. We'll provide a cloud sandbox with everything pre-installed, or you can bring your own setup with Claude Code CLI and an Anthropic API key (Pro tier).