Jia Huang

Claude Code Engineering in Practice

Packt Publishing Limited

From AI Coding Tool User to Agent System Builder

Session OverviewClaude Code is not just a command-line coding assistant — it is a programmable, extensible, and composable agent framework. In this 2-hour session, participants will learn how to engineer reliable AI agents by designing memory systems, delegating tasks to specialized sub-agents, and building reusable skill packages. Each concept is grounded in a hands-on case study drawn from production scenarios.Format: 2-hour live session (lecture + live demo + discussion)Target Audience: Developers, tech leads, and AI engineers with basic programming experience Module 1: Claude Code Architecture &amp; Memory System (30 min)1.1 Claude Code Tech Stack Overview (15 min)<ul><li>Architecture: CLI → Agent → Tool Chain → Extension Mechanisms</li></ul><ul><li>Core capability map: Memory / Sub-Agents / Skills / Hooks / MCP / Headless</li></ul><ul><li>How Claude Code differs from Cursor, Copilot, and other AI coding tools — it is a full agent framework, not just an autocomplete engine</li></ul>1.2 CLAUDE.md Memory System (15 min)<ul><li>The 3-tier memory hierarchy: Project-level / User-level / Global-level</li></ul><ul><li>Teaching Claude your project conventions, coding style, and team agreements</li></ul><ul><li>Case Study: How implementing CLAUDE.md raised code first-pass accuracy and team style consistency</li></ul> Module 2: Sub-Agents — Task Decomposition &amp; Collaboration (50 min)2.1 Core Concepts (15 min)<ul><li>Why split "one brain" into multiple "specialized roles"</li></ul><ul><li>Isolated execution, permission boundaries, and context management</li></ul><ul><li>Three canonical sub-agent patterns: Read-Only / Filter / Pipeline</li></ul>2.2 Case Study: The Bug-Fix Pipeline (35 min)A complete 3-agent pipeline that takes a bug from discovery to verified fix, demonstrating all three sub-agent patterns in one workflow:<ul><li>Locator Agent (Read-Only): Scans the codebase with Glob/Grep/Read to pinpoint the bug. No write permissions — cannot accidentally modify code.</li></ul><ul><li>Fixer Agent (Filter): Receives the locator's diagnosis, applies targeted edits with Read/Write/Edit. No execution permissions — cannot run anything.</li></ul><ul><li>Verifier Agent (Pipeline): Runs tests with Bash/Read to confirm the fix. No write permissions — cannot alter the fix.</li></ul>Discussion: How permission boundaries create reliability — a pattern transferable to any agent framework (LangChain, CrewAI, AutoGen). Module 3: Skills — Building Reusable Agent Capabilities (30 min)3.1 SKILL.md Structure &amp; Trigger Mechanism (10 min)<ul><li>The description field is a trigger, not documentation — writing it right is critical</li></ul><ul><li>Automatic discovery and on-demand loading: how Skills differ from Slash Commands</li></ul>3.2 Case Study: Progressive Disclosure Architecture — Financial Analysis Skill (20 min)A 3-layer architecture that minimizes token waste while maximizing capability:<ul><li>Index Layer (~100 tokens): Claude scans skill names and descriptions</li></ul><ul><li>Content Layer (&lt;5K tokens): SKILL.md instructions loaded only when matched</li></ul><ul><li>Appendix Layer (on demand): Referenced templates/standards loaded only when needed</li></ul>Live Demo: Installing the financial analysis skill and watching Claude automatically discover and apply it when asked to analyze a company's quarterly report. Wrap-Up &amp; Q&amp;A (10 min)<ul><li>Key takeaway: The gap between demo agents and production agents is engineering architecture, not smarter models</li></ul><ul><li>The four pillars of agent reliability: Memory, Isolation, Composability, Automation</li></ul><ul><li>Open Q&amp;A with participants</li></ul> What Participants Will Walk Away With<ul><li>A mental model for Claude Code as a programmable agent framework</li></ul><ul><li>Practical patterns (Memory, Sub-Agents, Skills) applicable to any agent system</li></ul><ul><li>Understanding of how permission boundaries and context isolation create reliability</li></ul><ul><li>A concrete architecture for building reusable, token-efficient agent capabilities</li></ul>

Our {communityGuidelinesLink} describe the sort of content we prohibit on Eventbrite. If you suspect an event may be in violation, you can report it to us so we can investigate.

To report other categories of prohibited or illegal content, submit {link}

Claude Code Engineering in Practice

More events from Packt Publishing Limited

Discover more events from Packt Publishing Limited, from Science & Tech to other experiences you might love.

Still looking for the right event?

Explore all online events to browse and filter by date, category, and more.

Claude Code Engineering in Practice

More events from Packt Publishing Limited

Discover more events from Packt Publishing Limited, from Science & Tech to other experiences you might love.

You might also like...

Browse more events with different dates, prices, and formats to find your next great experience.

Still looking for the right event?

Explore all online events to browse and filter by date, category, and more.