1-day Workshop · Practical AI training · Hands-on workshop
23/12/2026 · 09:00 – 17:30
👉 Save Your Seat HERE
Overview
AI coding agents have changed what a developer’s day looks like. Most developers have tried one. Far fewer use one well. The gap isn’t the tool. It’s knowing what the agent knows, how to ask it for work, and how to tell when its result is actually right.
This is a hands-on workshop built around Claude Code. Participants work in a prepared sample repository in the morning and in their own codebase in the afternoon, in whatever language they use. Every topic follows the same rhythm: a short concept, a live demonstration (including one where things go wrong), and a lab where each participant builds it themselves.
By the end of the day, every participant leaves with a working Claude Code setup for their own repository: a CLAUDE.md that teaches the agent their system, a skill for a task their team repeats, and a real task completed and verified.
What you will learn
- The vocabulary. What LLM, agent, harness, context window and software factory actually mean, and why the difference matters in practice.
- Prompting a coding agent. The anatomy of a work request, plan mode, and giving the agent a way to check its own work.
- What the agent knows about your system, and how to control it with CLAUDE.md, memory and context hygiene.
- Skills and commands. Teaching the agent, a recurring job once, so you stop repeating yourself.
- The map. Subagents, hooks, MCP and automation: what each one is for and when to reach for it.
Why this workshop
Buying licenses doesn’t change how a team works. Out of the box, an agent knows nothing about your codebase, your conventions or your domain. It produces code that looks right, compiles, and is wrong about something that matters. Developers lose trust within weeks, and adoption quietly dies.
The skills that close that gap aren’t about the model. They’re about the developer: asking precisely, supplying context deliberately, and making sure that a test, not a hopeful reading of a diff, says the work is done.
Target audience
- Developers with at least one year of professional experience, in any language.
- Basic familiarity with a chat assistant (ChatGPT, Claude.ai) is enough. No prior Claude Code experience is required.
Before the workshop
- A laptop with Claude Code installed and signed in. A short setup guide is sent a week ahead and takes about 20 minutes.
- A Claude account with Claude Code access, and git.
- Access to your own repository for the afternoon (recommended). If that isn’t possible, the whole day runs on the sample repository.
Workshop outline
- The vocabulary: model, harness, agent, factory
- Claude and Claude Code basics
- Prompting a coding agent
- Context and CLAUDE.md
- Skills and slash commands
- Putting it together: a real task in your own code
- The map: subagents, hooks, MCP and automation
- Wrap-up
Register for a Public Workshop
- Standard Price: 700 NIS + VAT per attendee
- Group Discount (3+ attendees): 630 NIS + VAT per attendee
► Contact Practical Software to register via this form
In-House Delivery
In-house workshops are tailored to your organization: your stack, tooling, security policies, and, where possible, your own codebase. Delivered in Hebrew or English.
► Contact Practical Software via this form to schedule an in-house workshop for your team.
* * *
Workshop Leader: Lior Friedman.
Lior is a Senior Software Architect and R&D Leader with 25+ years of experience building mission-critical systems and leading large-scale engineering transformations. Today, he focuses on helping R&D organizations adopt AI-enabled engineering practices-turning AI capabilities into practical, repeatable, and scalable workflows across coding, testing, and documentation.
* * *
