Claude Code, Codex and Cursor training for teams
Claude Code and Cursor training for your engineering team.
Your engineers already have the seats. A few hand the agent whole tasks, and the rest use it as autocomplete. I teach the whole team what the few do: AGENTS.md, skills and rules files, planning before the agent writes, and reviewing what it wrote. Whichever agent you pay for, in any language, usually on your own codebase.
For one engineering team. This is one half of the AI Engineering Workshop, and the call can add the other.
Same codebase, same agent, two different jobs.
What most engineers type
add pagination to the users endpoint
What the few type
Read AGENTS.md and the orders route first. Orders already has cursor pagination.
Add the same to GET /users. Plan it and wait for me before you write any code.
Add an integration test next to the others, run the test suite, and stop if anything outside the API and services folders needs to change.
The engineers who write the second one worked it out on your codebase, one rejected diff at a time. I teach the rest of the team to do the same.
What the coding-agents half covers.
The coding-agents half of the AI Engineering Workshop teaches an engineering team to hand real work to Claude Code, Codex, Cursor or Antigravity, usually on its own codebase.
The call picks which of these your team needs and in what order. Where there's a post on the subject, it's linked, free to read before you book.
Across Claude Code, Codex, Cursor and Antigravity
Context files, skills, planning and review carry across the four, so a team split across two tools still gets one workshop. Where a tool has a file of its own, we work with the one your engineers use.
Context files
An AGENTS.md at the root of the repository, plus CLAUDE.md or the tool's own rules files where your engineers use them: the framework, where things live, how the tests run and what the agent must not touch. With that in place the agent follows your conventions on the first try instead of guessing at them from three greps.
Skills
SKILL.md files for the procedures your team repeats, and knowing when a skill is the wrong tool and one line in AGENTS.md does the job, since a skill only helps once the agent decides to open it.
When to write an agent skillPlanning before the agent writes
Asking for a plan, reading it and correcting it before any code exists, the way the second prompt above does, and setting up the environment and the first passing test before the feature work starts.
Reviewing diffs and tests
Reading what the agent changed, asking for the test that proves it, and running the suite before accepting anything. A check the agent doesn't control decides when the work is done.
Deciding which tasks to hand over
Which work goes to the agent whole, which goes in pieces and which stays with an engineer, decided task by task on your own backlog, so nobody stops understanding what ships.
What is loop engineeringSandboxing and permissions on a real repository
What the agent may run, read and change: the permission settings, the sandbox the tool already ships with, and the commands that should wait for a person to approve them.
Your engineers leave with the code they wrote.
I explain each moving part and they write it, on content shaped around what they are building. No pre-work for attendees. Remote or on-site.
A 30-minute call
What your team is building, which coding agents they use, and which half of the workshop they need. Often it is both.
Shaped around your team
I put the content together for your team and the work in front of them. How long it runs depends on the scope we agree.
The workshop
Remote or on-site. I explain each part, and your engineers do it as we go, on their own code with their own coding agent or by writing the agent themselves.
Two halves, one workshop.
The call decides which half your team needs, or both, and in what order.
Working with coding agents
Cursor, Claude Code, Codex or Antigravity, whichever your team uses, in any language, usually on your own codebase. Giving the agent the context it needs with AGENTS.md, skills and rules files, planning before it writes, reviewing what it wrote, and deciding which work to hand it in the first place.
Building agents
Agents, agentic workflows, harness engineering and the other moving parts, with your engineers writing the code as we go. The syllabus is the free book, and the team needs basic JavaScript or TypeScript.
Who is in the room
One engineering team. For working with coding agents, any language. For building agents, basic JavaScript or TypeScript and a basic understanding of large language models. No machine learning background for either.
Where the wider problem is that most of an organisation never got past asking these tools questions, that's different work: the Power User Gap.
What they leave able to do
Depending on what the call settles, they hand real work to the coding agent they already have and know how to check what comes back, or they understand how agents are built because they wrote the moving parts themselves. Often both.
Before you book
- Do engineers need to use the same coding agent?
- No. Context files, skills, planning and review carry across Claude Code, Codex, Cursor and Antigravity, so a team split across two tools still gets one workshop. Where a tool has its own file, such as CLAUDE.md or a rules file, we cover the one your engineers use.
- Does it work if our codebase isn't JavaScript?
- Yes. The coding-agents half works in any language, and usually runs on your own codebase. JavaScript or TypeScript only matters for the building-agents half.
- Is it for teams already using Claude Code, or ones about to roll it out?
- Both. A team that already has the seats starts from how it uses them now. A team about to roll one out starts with the context files and the permissions, before the agent gets its first real task.
- Is it for the whole team, or only the engineers already getting results?
- The whole team. The engineers who already hand the agent whole tasks are welcome, and the teaching is aimed at the rest.
- What do attendees need to prepare?
- Nothing. There is no pre-work. The call decides what goes in.
- What does it cost?
- It is quoted against the scope we agree on the call. There are no individual seats and no published prices.
Questions about model providers, recording and the teaching itself are answered on the AI Engineering Workshop page.
What people say about the teaching.
Our audience has been tremendously impacted with Tamas's educational content throughout the past years, leaving with practical insights & knowledge ready to be applied on Monday!
J. GerardsFrontend RevolutionTamas isn't just a well-prepared professional, he's also someone who offers insightful perspectives on the present and future of AI. I really enjoyed exchanging experiences with him and picking up some of his know-how along the way.
G. GalassiFreelance developer
Who you'd be working with
I use coding agents every day and keep building agents and harnesses to stay sharp on AI engineering, and I write about it in the open.
Google Developer Expert in Web Technologies, Agentic AI Foundation ambassador, MBA. 25 years across software engineering, developer relations and technical training, and 151 talks and workshops in 35 countries.
More about meStart with a 30-minute call
Tell me what your team is building and what they need to learn. It's one quote, for one engineering team and the scope we agree on the call, one half or both, with no individual seats, so the figure you forward is the whole workshop.