Workshop teaches developers to automate software tasks with Claude Code

From code lines to full tasks

Artificial intelligence in software development is moving past the stage of merely suggesting the next line of code. Modern tools are increasingly capable of handling entire development tasks autonomously. This shift requires a different approach to how developers interact with these systems.

The upcoming iX-Workshop focuses on Claude Code, which functions as an agentic tool. Unlike simple chat interfaces, an agentic tool can perform a series of actions to reach a goal. This allows developers to significantly speed up their workflows by delegating larger blocks of work to the AI.

As heise.de reported, the goal is to help developers spend less time on manual coding and more time on high-level logic. By using the right prompting strategies and context management, teams can increase the efficiency of their entire development projects.

Mastering agentic workflows

The workshop provides a deep dive into how to manage Claude Code for complex requirements. This includes everything from the initial architecture planning to the final implementation of features. Participants will also learn how to handle testing and refactoring, which are often the most time-consuming parts of the cycle.

A key part of the training involves teaching developers how to customize the AI. Users can adapt Claude to follow specific rules, coding standards, and unique skills required by their specific organization. This ensures that the AI-generated output aligns with the existing codebase and team preferences.

Technical depth is a major component of the curriculum. The session covers advanced features such as subagents, which allow the AI to break down tasks into smaller pieces. It also explores worktrees and the connection of external tools through MCP servers to automate and run tasks in parallel.

Practical learning with TypeScript

To ensure the skills are applicable, the workshop uses a continuous TypeScript project as the primary teaching vehicle. Developers will work through the material step-by-step, applying the methods directly to a real project. This hands-on approach helps build a practical sense of how agentic development feels in a live environment.

By working through a consistent project, participants can see how the AI handles different stages of a software lifecycle. This helps them identify where the AI excels and where its limitations lie. Understanding these boundaries is essential for integrating AI into a professional production environment.

The training is led by Rainer Stropek, who brings experience in software architecture and cloud computing. He has worked on SaaS solutions and focuses on the productive use of AI-assisted development tools. His experience with real-world projects helps ground the workshop in practical reality.

Who is this for?

This workshop is specifically designed for experienced software developers. It is not intended for beginners who are just starting to learn how to code. Instead, it targets those who already have some experience with AI-assisted development but want to do more with it.

The target audience includes developers who want to make their existing processes more efficient. They are looking to move from manual prompting to a more sophisticated use of agentic AI. The small group size is intended to provide room for individual questions and the discussion of specific use cases.

Participants can bring their own challenges to the table to get concrete impulses for their daily work. The goal is for every attendee to leave with a better understanding of how to weave these agents into their current professional routines. This focus on integration makes it highly relevant for teams looking to scale their output.

The future of development

The move toward agentic tools like Claude Code represents a significant change in the developer experience. Instead of being a passive recipient of code suggestions, the developer becomes a supervisor of an AI agent. This requires a shift in mindset toward managing context and defining clear goals for the machine.

As tools become more capable of handling subtasks and connecting to external servers, the complexity of what an AI can do in a single session grows. Developers who master these skills will be able to handle larger projects with fewer manual interventions. This could lead to faster release cycles and more time spent on creative problem solving.

The workshop highlights that while the technology is powerful, it still requires human oversight. Knowing where the AI hits a wall and how to guide it back on track is the primary skill being taught. This balance of human direction and machine execution is the core of modern software engineering.