Course
The Prompting Playbook: Debug a Production Prompt
6 lessons
Start free — no account, no card.
One whole course is free. After that, one payment of $10 opens all 15 courses in the catalogue today. What’s included →
An interactive walkthrough of "The Prompting Playbook", a public breakout talk at the Code with Claude conference by Margot van Laar, Applied AI Engineer at Anthropic in London. Follow her live debugging of a messy customer-support prompt for a fictional telco: why a prompt that worked breaks on a new model, how to build an eval suite of control, edge and capability-boundary cases, how to strip pasted-website cruft and an untrue persona, how XML tags separate role from policy from tone, and how an output contract plus a stop sequence pin down the response — then isolate a single failing case and find the conflicting instruction behind it. Source: Margot van Laar, "The Prompting Playbook", Code with Claude.
What you’ll learn
1.Stop prompting on vibes
A production prompt regresses on a new model. Before touching a word of it, work out what kind of problem you have — and build the eval suite that lets you tell whether any change actually helped.
2 lessons
2.Clean it before you fix it
General hygiene first, specific failures second. Strip the pasted website furniture and the untrue persona, then carve the remaining wall of text into tagged sections — and watch the eval move without a single new instruction.
2 lessons
3.Control output, then target failures
Pin down where the answer lives with an output contract and a stop sequence, then isolate a single failing case and find the instruction that caused it. The talk's recording ends inside this last fix, and so does this module.
2 lessons
How this course works
This is an interactive course, not a video playlist. Lessons mix short explanations with hands-on activities — quizzes, drag-and-drop exercises, branching scenarios and step-ordering challenges — and you have to complete each activity before moving on, so you practise rather than skim. Your progress is saved as you go, so you can stop after any lesson and pick up where you left off.
Topics covered
- prompt-engineering
- evals
- llm
- production
- claude
- anthropic
- debugging
- interactive