Agentic AI Enablement: Claude Code and ChatGPT Codex.
The compact entry into agentic development with Claude Code and ChatGPT Codex. In four hours the coach walks you through the fundamentals of both tools and builds a sample project before your eyes: you experience how a coding agent plans, builds and tests, and you see every step in the session log.
Good reasons to get started.
From chatbot to agent
With a chatbot the work ends in text to copy. An agent reads the files in the project, runs tests and talks to Git. The course demonstrates this difference on real workflows, including the points where you step in.
Two tools, one way of working
Claude Code and ChatGPT Codex share the rules file, approval modes and parallel workers. The course maps the vocabulary line by line, so you can keep using the subscription you already pay for.
The coach builds, you see every step
The sample project Mini-Markt takes shape during the course on the shared screen: briefing, plan, tests, two features, review and the merged pull request. The coach conducts all exercises; you follow every step in the session log and ask questions along the way, without setting up anything yourself.
Costs and control in numbers
Models, subscriptions and token prices of both vendors in direct comparison, plus the permission levels from plan mode to auto. Afterwards you can put a number on what a working day with the agent costs, and justify which approvals it gets.
How the course is structured.
The map: two agents, one way of working
The Anthropic and OpenAI platforms at a glance: models, subscriptions, prices. Plus a look at the release cadence that turns chasing the current leader into a Sisyphean task.
- Place the model families of both vendors (Haiku to Fable, Luna to Sol)
- Estimate subscription and API costs realistically
- Vocabulary mapping: AGENTS.md and CLAUDE.md, cloud tasks and subagents
- Get to know Cowork and Design as Claude extras
Setting up and steering Claude Code
A tour of the desktop app, access through the subscription and control over every step: modes, diff view and the GitHub connection in two steps.
- The interface and the twelve everyday moves
- Choose permission levels from plan to auto with confidence
- Connect GitHub, with your own permissions
- Understand the effort dial and limit consumption
Feature development with the agent
The centrepiece: the coach builds the sample project Mini-Markt from a briefing and extends it with features in the rhythm of goal, plan, diff, test. You follow every decision on the shared screen.
- Create and sharpen CLAUDE.md as the project memory
- Phrase tasks as outcomes, read the plan, approve deliberately
- Tests, review and pull request in the agentic flow
- A red test as information the agent keeps working with
Toolbox and outlook
Slash commands for everyday work, subagents, hooks and MCP. To close, a look at Cowork and Design with the handover of a draft to Code.
- The most important slash commands, each with its Codex counterpart
- Hooks: what must never break is enforced by the system
- A custom command from a Markdown file
- Design draft with handover to Claude Code
What you can do after the course.
What it costs.
What is included
- Live introduction with a demonstrated sample project
- Slides and cheat sheet for all commands
- Access to the online course material
- Confirmation of attendance
Discounts
Early-bird discount 10% when booking more than 30 days before the date. Group discount 10% from 5 participants registered together.
In-house training?
For an on-site learning environment and a format that you can run repeatedly for your colleagues, we deliver this workshop in-house as well. Reach out for a tailored offer.
Frequently asked.
Do I need my own subscription?
Not for this variant. The coach conducts all exercises himself, all you need is a screen. If you want to click along in parallel, you need your own Pro subscription; it only becomes mandatory in the Deep Dive.
I already work with Codex. What does the course give me?
The complete Claude Code craft plus the translation table back into your Codex world. The concepts of both tools largely overlap, and the course maps the vocabulary line by line; some moves, like rewinding a session with rewind, you will meet here for the first time.
What separates this variant from the Deep Dive?
The distribution of roles. Here the coach conducts all exercises himself, you watch and ask questions in between; your own setup and subscription are not required. In the Deep Dive it is the other way round: every exercise runs on your machine, from setup through CLAUDE.md to the merged pull request, with the coach at your side.
Is attendance documented?
Yes, all participants receive a confirmation of attendance listing the covered content. It works for internal records and HR files.
Work agentically, with the tool of your choice.
Four hours, a project built live by the coach up to the merged pull request, plus the cheat sheet for all commands to rework on your own machine.
Maybe a different pillar fits your situation better.
Quality Consulting
Strategie, Methodik, Frameworks für belastbare Qualität. Audits, Konzepte, AI-Compliance.
→Quality Services
Operative Test-Manpower, Interim-Testmanagement und Vermittlung aus dem Fachnetzwerk.
→Quality Education
Workshops, Schulungen und 1:1-Coaching für Test-, Projekt- und KI-Compliance-Themen.
→CT Map
Übersicht aller drei QCT-Säulen mit Wegweiser zu deinem passenden Einstiegspunkt.
→