Training course · Claude Code · Game Development

Claude Code for Game Developers

Agentic coding with Claude Code in the terminal and IDE: CLAUDE.md, subagents, hooks, MCP, CI pipelines and secure permissions.

Duration
2 days (14 hrs)
Format
Online or on-site
Language
German or English
Participants
open dates or exclusive for your team
Certificate
Certificate of attendance

Overview

“Claude Code for Game Developers” is designed for roles such as Gameplay programmers working in Unity, Unreal Engine or Godot, Indie developers and small studios and Technical artists and tools programmers and shows step by step how to use Claude Code productively, safely and transparently in Game Development – from the fundamentals to advanced workflows.

Claude Code is Anthropic’s agentic coding assistant, working directly in your terminal and in popular IDEs. Rather than completing individual lines, it reads your repository, plans changes across multiple files, runs commands, executes tests and prepares commits. Over two days you will learn to use it with deliberate control: clear project instructions in CLAUDE.md, careful context management, and a workflow in which you set the direction as a developer and can trace every change the agent makes back to a clear intention.

The course is built around real engineering work: diagnosing bugs, delivering features, refactoring legacy code and filling gaps in test coverage. You will configure subagents for specialised tasks, enforce quality checks with hooks, connect external systems through the Model Context Protocol (MCP) and capture team knowledge as reusable skills. Finally, you will run Claude Code headless in CI pipelines and define permissions, sandboxing and approval rules so that its use remains defensible even in security-sensitive environments.

Game development combines real-time performance, complex state logic and creative iteration like few other fields. AI coding assistants and agents help teams working in Unity, Unreal Engine and Godot prototype gameplay systems faster, automate editor tooling, write shaders, analyse performance bottlenecks and refactor long-lived codebases. Small indie teams gain time for game feel and content, while larger studios make tooling and pipelines more efficient. The domain also raises particular questions: frame budgets, garbage collection spikes, multiplayer synchronisation and the rights to generated content. In this course you learn to use AI purposefully within your engine workflow, from C# scripts in Unity and C++ and Blueprints in Unreal to GDScript in Godot, while keeping quality and playability firmly in view.

The course is delivered live – either online in a virtual classroom or on-site in our training rooms in Munich, Salzburg, Berlin, Augsburg, Hanover, Bremen, Stuttgart and Kaiserslautern. We can also run it exclusively for your team.

Who is this course for?

  • Gameplay programmers working in Unity, Unreal Engine or Godot
  • Indie developers and small studios
  • Technical artists and tools programmers
  • Engine and performance specialists
  • Developers of serious games, simulations and interactive applications

Typical challenges we address

  • Intense iteration pressure when prototyping new game mechanics
  • Performance issues such as frame drops, memory spikes and long load times on target platforms
  • Tightly coupled legacy gameplay code that makes extensions difficult
  • Repetitive editor and pipeline tasks that eat into creative time

What you will be able to do

  • You use Claude Code for multi-step development tasks while keeping full control over every change.
  • You maintain a CLAUDE.md that effectively conveys project conventions, commands and architectural guidance to the agent.
  • You extend Claude Code with subagents, hooks, MCP servers and skills tailored to your team.
  • You integrate Claude Code into Git workflows and CI pipelines with traceable, test-backed results.
  • Prototype game mechanics and editor tools much faster with AI while staying in control of the architecture.
  • Analyse and optimise performance issues with AI support based on real profiler data.
  • Apply AI agents in a way that suits Unity, Unreal Engine or Godot and understand the legal questions around generated content.

Course outline

Day 1

Module 1 How Claude Code works
  • Installation, sign-in and use in the terminal, VS Code and JetBrains IDEs
  • Understanding the agentic loop: read, plan, act, verify
  • Using plan mode to agree on an approach before any code changes
  • Writing effective task briefs with goal, context and acceptance criteria
Module 2 Project context with CLAUDE.md and context management
  • Structuring CLAUDE.md at project, directory and user level
  • Documenting build, test and style conventions so the agent follows them reliably
  • Managing the context window with targeted file references, compaction and fresh sessions
  • Resuming sessions and breaking long tasks into manageable stages
Module 3 Extending Claude Code with subagents, hooks, MCP and skills
  • Defining subagents for code review, research or test generation
  • Using hooks to enforce formatting, linting or checks before and after tool calls
  • Connecting MCP servers for issue trackers, databases or internal documentation
  • Creating custom slash commands and skills for recurring team workflows
Module 4 Git workflows, testing and quality assurance
  • Creating branches, meaningful commits and pull requests with Claude Code
  • Test-first work: have tests written before the implementation
  • Reviewing diffs systematically and rejecting or refining changes
  • Organising parallel work with Git worktrees and multiple sessions

Day 2

Module 5 Headless operation, CI and security
  • Running Claude Code non-interactively in scripts and CI pipelines
  • Automated reviews and issue handling in GitHub or GitLab workflows
  • Configuring permission modes, allow and deny rules, and sandboxing
  • Handling secrets, prompt injection risks and organisation-wide policies
Module 6 Practice lab: Game Development
  • Prototyping gameplay mechanics
  • Automating editor tools
  • Developing shaders
  • Analysing performance
  • Refactoring gameplay code
  • Implementing multiplayer logic
Module 7 Data protection, quality & transfer to daily work
  • Unreleased games, engine source code and assets are valuable intellectual property: you learn to respect the licence terms of engines and AI services, clarify rights to generated code and content, understand platform rules on disclosing AI-generated material and process player data in line with the GDPR.
  • Ground rules for safe AI use in your team
  • Checking results: facts, sources, tests and the four-eyes principle
  • Personal action plan for the next 30 days
  • Open questions, outlook and resources for further learning

Practical scenarios from your field

Prototyping gameplay mechanics

Generate working scripts for movement, combat systems, inventories or dialogue systems from a game description, exposing parameters that designers can fine-tune directly in the editor.

Automating editor tools

Write custom editor windows, import scripts, asset and level validation and batch operations that reliably handle recurring tasks in Unity, Unreal or Godot. Asset errors surface earlier.

Developing shaders

Describe visual effects and have them implemented in HLSL, Shader Graph, Unreal materials or Godot shading language, with explanations of the maths and notes on mobile performance.

Analysing performance

Interpret profiler output, identify allocations in the update loop and expensive physics or rendering calls, and propose optimisations such as object pooling, batching or data-oriented design.

Refactoring gameplay code

Turn monolithic manager classes into component-based, event-driven structures, model state machines cleanly and add tests for gameplay-critical logic. Game feel is preserved while new features become easier to add.

Implementing multiplayer logic

Design network synchronisation, client-side prediction, authoritative server logic and remote procedure calls, analyse typical causes of desyncs and build anti-cheat considerations into input validation. Latency and packet loss can be simulated.

Possible hands-on projects

Level validator

An editor tool that checks scenes for missing colliders, wrong layers, unused assets and naming conventions and produces a report with clickable links.

Dialogue editor

A small tool that lets writers build branching dialogue visually and export it as JSON for the engine to load at runtime.

Playable mini prototype

A complete small prototype, such as an arena shooter or puzzle level, with controls, scoring and simple enemy behaviour. It helps validate game feel early.

Prerequisites

You have hands-on programming experience in at least one language and are comfortable with Git and the command line. You have programming experience in C#, C++ or GDScript and have worked with a game engine such as Unity, Unreal Engine or Godot.

Technical requirements

Your own laptop with admin rights, Git, a Node.js runtime and an up-to-date editor; a Claude subscription or Anthropic API access is required for the exercises.

Data protection & compliance

Unreleased games, engine source code and assets are valuable intellectual property: you learn to respect the licence terms of engines and AI services, clarify rights to generated code and content, understand platform rules on disclosing AI-generated material and process player data in line with the GDPR.

Suggested dates

All dates are suggestions – your preferred date is always possible.

  • Tue, 24 November 2026 – Wed, 25 November 2026 Online Request
  • Wed, 2 December 2026 – Thu, 3 December 2026 Hanover Request
  • Tue, 15 December 2026 – Wed, 16 December 2026 Online Request
  • Mon, 18 January 2027 – Tue, 19 January 2027 Munich Request
  • Wed, 3 February 2027 – Thu, 4 February 2027 Online Request
  • Tue, 16 February 2027 – Wed, 17 February 2027 Stuttgart Request

Frequently asked questions

How is Claude Code different from traditional code completion?

Code completion suggests the next few lines, whereas Claude Code acts as an agent that handles whole tasks: it searches the repository, plans changes across several files, runs tests and commands, and corrects itself based on the results. You define the goal and review the outcome instead of steering every single line yourself.

Is Claude Code secure enough for enterprise use?

By default, Claude Code asks for approval before editing files or running commands. Permission rules, sandboxing, managed settings and hooks let you define precisely what the agent may do. During the course you learn to configure these mechanisms to match your security and data protection requirements and to assess risks such as prompt injection.

Does Claude Code work with my language and tech stack?

Yes. Claude Code is language-agnostic because it works with files, terminal commands and your existing build and test tooling. Whether you use Python, Java, TypeScript, Go, SAP-related code or infrastructure as code, what matters most is a well-written CLAUDE.md describing your commands and conventions. Each course variant adds the relevant domain focus.

Which engine is the course designed for?

The course is engine-agnostic and covers Unity with C#, Unreal Engine with C++ and Blueprints, and Godot with GDScript or C#. You can complete the exercises in your preferred engine. For each, we show how to provide project structure, engine APIs and conventions as context so the AI assistant delivers fitting solutions.

Can AI-generated code and content be used in commercial games?

For code this is generally possible, provided the AI service’s terms allow it and you plan for review and licence checks. For artwork, audio and text, the legal situation and individual storefronts’ disclosure requirements need attention. We give a practical overview in the course, though it does not replace legal advice.

How much does “Claude Code for Game Developers” cost?

The course costs a flat €949 plus VAT per participant – whether online or on-site. This includes 2 training day(s) with a live trainer, course materials, a lab environment and a certificate of attendance.

Can I suggest my own date?

Yes. Our suggested dates are just a guide. In the request form you can always describe your preferred date or period in free text – we check availability and confirm within one working day.

Is the course online or on-site?

Both are possible: live online in a virtual classroom or on-site in Munich, Salzburg, Berlin, Augsburg, Hanover, Bremen, Stuttgart or Kaiserslautern. We can also run the course exclusively for your team.

More courses for Game Developers

More Claude Code courses

All 77 Claude Code courses

Claude Code · MunichClaude Code · SalzburgClaude Code · BerlinClaude Code · AugsburgClaude Code · HanoverClaude Code · BremenClaude Code · StuttgartClaude Code · KaiserslauternClaude Code · Online