Prompt Engineering

This course teaches you to talk to models like a developer — without writing a line of code: clarity and specificity, context and examples, role and tone, structure and structured output, step-by-step reasoning, and how to diagnose a bad output and fix the prompt. You'll also meet the concepts behind the scenes — system vs user, tool use, and the generation controls (temperature, length, stop sequence) — all explained conceptually.

Lesson 1: What a prompt is, and why wording decides the answer
Lesson 2: Clear vs vague — say exactly what you want
Lesson 3: Specificity — format, count, and length
Lesson 4: Context and delimiting the input
Lesson 5: Positive vs negative instructions, and giving the model an 'out'
Lesson 6: Examples (few-shot) — teach by showing
Lesson 7: Role, persona, and tone
Lesson 8: Structure and format — asking for output that's easy to read and use
Lesson 9: Step-by-step reasoning (chain-of-thought)
Lesson 10: Decomposing a complex task into steps
Lesson 11: Reading and diagnosing a bad output
Lesson 12: Improving a prompt through iteration
Lesson 13: Prompt injection and safety — in plain terms
Lesson 14: system vs user — the two layers of a request
Lesson 15: tool use — giving Claude tools
Lesson 16: generation controls — the dials a developer tunes