Course · Training · Workshop

Advanced Claude Code

For everyone already using Claude Code daily: context architecture, models and modes, hooks and sandboxing, your own skills, plugins and MCP, subagents and parallel agents, headless in the pipeline and rollout across the team.

Claude Code is far more than a coding assistant in the terminal: it is an extensible agent platform with skills, subagents, hooks, plugins, MCP, headless mode and remote sessions. This follow-up course is for developers who already use Claude Code every day and now want to get everything out of it. We look under the hood - what enters the context when, what each feature costs - and build a clean context architecture for large codebases and teams on top of that. Participants choose models, effort and permission modes deliberately, enforce rules with hooks and sandboxing instead of only describing them, and write their own skills, subagents and MCP integrations, which they distribute as a plugin through a team marketplace. They run several agents in parallel in worktrees, orchestrate fan-out workflows with independent verification, and make agent output verifiable with tests and agentic review. Finally they bring Claude Code headless into the CI/CD pipeline, use remote sessions and scheduled routines, and keep cost, monitoring and governance under control across the team. All hands-on, on a realistic application in your personal letsboot labmachine.

In-house course

Tailored for your team.

Request

No suitable date?

Request a new public date.

Request a date

Trainers

Christian Wörz

Instructor

Vincent Grimmeisen

Instructor

Andreas Deicha

Instructor

Jonas Felix

Instructor, Co-Founder

What participants say

The “AI Assisted Engineering On-Boarding” course was an outstanding experience. It provided a clear, practical, and inspiring introduction to how AI can support modern engineering work. The content was well structured, easy to follow, and highly relevant to real day-to-day challenges. What stood out most was the strong balance between theory and hands-on application. The course not only explained the concepts clearly, but also showed how to use AI tools effectively and responsibly in actual engineering scenarios. It made complex topics feel accessible and immediately useful. I especially appreciated the engaging format, the high quality of the materials, and the clear expertise of the instructors. The course created a motivating learning environment and gave me valuable confidence in applying AI-assisted approaches in my work. Overall, I can highly recommend “AI Assisted Engineering On-Boarding” to anyone who wants a practical, forward-looking, and well-designed introduction to AI in engineering. It is an excellent course that delivers real value from start to finish.

Martin
Leiter IT Entwicklung Platform & Services

The AI Assisted Engineering On-Boarding course was very practical and easy to follow and awesome! I especially appreciated the hands-on approach and the focus on real use cases. It gave me a clear understanding of how to work effectively with Claude Code in a development context. After the course, I felt much more confident getting started and applying it in daily work.

Alexander
Frontend Web Developer

The Agentic Product Engineering course gave me a completely new perspective on modern product development with AI agents. From the fundamentals to advanced workflows, everything was explained clearly and presented in a way that was easy to understand and apply. The concepts and techniques I learned could be integrated directly into my daily work. I was impressed and engaged from start to finish. Highly recommended for anyone looking to leverage AI agents in product development.

Daniel
Chief Product Officer

The day was both intense and practical, with a range of exercises and challenges designed to provide valuable insights into agentic AI systems and modern product engineering.

Pascal
Head Digital Transformation

The course fully met all my expectations and provided a smooth introduction to the subject. With the help of the instructor and the course materials, I was able to easily overcome any initial hurdles, allowing me to get started using the agents as quickly as possible.

Daniel
Senior Software Engineer

I really enjoyed the lessons and learned a lot of new stuff on this course. It teaches not only fundamentals but also how to be prepared for your next challenges. The tooling that is used is very sturdy, works very well in the lab setup and is very diverse, you can pick what you like in your projects. And most importantly: I did not write any piece of code myself, the agents did everything!

Andrei-Flavius
Software Engineer I

* Voices from this or related courses.

These customers booked courses in the same topic cluster.More customers →

Content

The course consists of the following topics and is tailored to your codebase, toolchain and focus areas for in-house runs. Unlike the tool-agnostic course Advanced Agentic Software Engineering, this course goes deep into the Claude Code platform itself.

– Claude Code under the hood:

Agent loop, context window and what loads when
How much context and tokens each feature costs
Sessions, resume, compaction and checkpoints with rewind
Hands-on: analyse a session and halve its context usage – Context architecture for large codebases and teams:
CLAUDE.md hierarchy, imports and path-scoped rules
Auto memory, AGENTS.md and shared team conventions
Monorepos and multiple repositories without an overloaded context window
Code intelligence with LSP plugins instead of blind grep
Hands-on: build a context setup for an unknown multi-repo project – Models, effort and modes:
Model choice, effort levels and fast mode per task
Using plan mode, permission modes and auto mode deliberately
Setting a verifiable goal with /goal instead of a prompt
Hands-on: compare the same task across models and modes – Control and security:
Permissions, sandboxing and managed settings
Hooks: events, command, prompt and agent hooks as guardrails
Reliably protecting sensitive files, secrets and generated code
Hands-on: enforce team guardrails with hooks instead of describing them in CLAUDE.md – Skills, plugins and marketplaces:
Writing skills with scripts, frontmatter and fork context
Reviewing third-party skills before installing them
Bundling skills, subagents, hooks and MCP into a plugin
Hands-on: build a team plugin and distribute it through your own marketplace – Using MCP deliberately:
Tool search, remote servers and OAuth
Connecting documentation, browser and tickets (Context7, Playwright, Chrome)
Assessing data flows, permissions and risks of MCP servers
Hands-on: build a small MCP server of your own for an internal system – Subagents, parallel agents and workflows:
Custom subagents, forks and background agents
Working in parallel with Git worktrees and agent teams
Fan-out workflows with independent, adversarial verification
Hands-on: split a migration across several agents and merge the results – Verification, testing and review:
Verification loops instead of “looks plausible”
Test-driven development with Claude and browser tests with Playwright
Agentic code review locally and in the merge request
Hands-on: secure a risky change with tests and multi-agent review – Claude Code headless and in CI/CD:
claude -p, structured output and budget limits
Claude Code in GitLab CI/CD and GitHub Actions: review, fixes, release notes
Outlook Agent SDK: your own agents built on Claude Code
Hands-on: integrate a merge request reviewer into the pipeline – Remote, asynchronous and in the team:
Claude Code on the web, remote control and mobile sessions
Routines, scheduled tasks and loops for recurring work
Cost, monitoring with OpenTelemetry and enterprise operation (Bedrock, Vertex, LLM gateway)
Hands-on: sketch a rollout plan for Claude Code in your own team

All concepts are centred on Claude Code. Where it helps, we show what transfers to Codex, OpenCode or GitHub Copilot.

The actual course content may differ from the above depending on the trainer, delivery, duration and the composition of participants.

Goal

After this course you master Claude Code as a platform instead of a chat window. You understand how context, models and modes interact and steer quality, speed and cost on purpose. You build reusable extensions - skills, subagents, hooks, MCP integrations - and distribute them as a plugin in your team. You use parallel agents and multi-agent workflows safely, verify their results systematically and integrate Claude Code into your CI/CD pipeline. And you know how to roll Claude Code out across your team in a controlled way with permissions, managed settings and monitoring.

Duration

2 Days

Format

The course is hands-on throughout: every short theory and demo block is followed by a practical exercise. Everyone works in their own letsboot labmachine with Claude Code in the terminal, in code-server and in the browser on the same realistic application. Every chapter starts independently from a defined state, so nobody falls behind. Participants take the skills, hooks and plugins they build straight into their daily work.

Target group

The training is aimed at experienced software developers, tech leads, platform engineers and architects who already use Claude Code regularly and want to become power users - or who set up and support Claude Code for their whole team.

Prerequisites

Solid programming experience, familiarity with Git, merge requests and CI/CD pipelines, and several weeks of hands-on experience with Claude Code (or a comparable coding agent). The course Claude Code for Enterprise Engineering is a good basis but not a prerequisite.

Preparation

Before the course every participant receives a questionnaire on experience and expectations plus access to their personal GitLab project and labmachine. Claude Code, the training application and all tools are preinstalled; a browser is enough. We provide model access during the course. After the course the labmachine environment remains available for your own further work.

Request this course in-house

Request a public date

No suitable public date? Register without obligation — once there is enough interest we schedule a new public date and let you know first.

More about Claude Code

Claude Code is Anthropic's agentic coding tool. It runs in the terminal and the IDE, understands whole codebases and carries out tasks on its own. With skills, subagents, hooks, plugins and the Model Context Protocol (MCP) it can be tailored to team and company needs, and in headless mode (`claude -p`) it runs directly in CI/CD pipelines.

Further resources:

History

Claude Code was first released by Anthropic as a terminal agent and has grown a lot since: skills, subagents, hooks, plugins and marketplaces turn the assistant into an extensible platform. The open Model Context Protocol connects any tool or data source.

This shifts the question from "What can the agent do?" to "How do we steer, verify and scale agents in a team?" - which is exactly what this course is about.