Module 42Stage: Create90 min

Prompt Engineering

Write a prompt with the four core components: context, task, constraints, example.

Jump to lesson:
01Prompt Structure02Context03Constraints04Examples05Prompt Chains
1
Lesson 1

Prompt Structure

Objective

Write a prompt with the four core components: context, task, constraints, example.

Teach
A strong prompt gives the AI context (who/what), a clear task, constraints (tone, length, format), and ideally an example of the desired output — vague prompts produce vague results.
Real-World Example
"Write a caption" (weak) vs. "Write a 3-line Instagram caption for a Lagos meal-prep brand, casual tone, ending with a specific CTA to DM 'MENU'" (strong).
Common Mistake
Writing one-line, context-free prompts and being surprised by generic output.
Practical Exercise

Rewrite one weak prompt using context, task, constraints and an example.

Lesson Checkpoint

A strong prompt typically includes

2
Lesson 2

Context

Objective

Give AI enough background to produce relevant, specific output.

Teach
Context means telling the AI who the brand is, who the audience is, and what's already been tried — exactly what KoraSpace's Brand Brain (Section 7.1) automates so you don't re-explain it every time.
Real-World Example
"Our audience is first-time founders in Lagos, skeptical of hype" changes the output significantly compared to no audience context at all.
Common Mistake
Assuming the AI "remembers" context from a previous, separate conversation or task.
Practical Exercise

Write the context block you'd include in a prompt for your brand.

Lesson Checkpoint

Providing context in a prompt mainly helps by

3
Lesson 3

Constraints

Objective

Use constraints to control tone, length and format precisely.

Teach
Constraints (word count, tone words, format requirements, things to avoid) narrow the AI's output toward what you actually need, reducing editing time afterward.
Real-World Example
"Under 150 characters, no emojis, no exclamation marks" produces a very different, more usable draft than an unconstrained prompt.
Common Mistake
Leaving constraints out and then being frustrated the output needs heavy editing.
Practical Exercise

Add 3 specific constraints to a prompt you'd normally leave open-ended.

Lesson Checkpoint

Constraints in a prompt mainly help by

4
Lesson 4

Examples

Objective

Use a real example to anchor tone and style precisely.

Teach
Including a short example of the desired output (or a past post that worked) gives the AI a concrete target that's more reliable than describing tone in the abstract.
Real-World Example
Pasting a previous high-performing caption as a style reference produces more consistent voice-matching than just saying "casual and direct."
Common Mistake
Describing tone only in adjectives ("fun," "professional") without giving a concrete example to anchor it.
Practical Exercise

Add one real example of your brand voice to a prompt template you use often.

Lesson Checkpoint

Including an example in a prompt mainly helps with

5
Lesson 5

Prompt Chains

Objective

Break a complex task into a sequence of smaller, connected prompts.

Teach
Instead of one giant prompt trying to do everything, break work into steps — research, then draft, then refine — similar to KoraSpace's own "Composing Post" pipeline (Section 5.1), which explicitly reflects on its own output before returning it.
Real-World Example
Prompt 1: research the niche and audience. Prompt 2: draft based on that research. Prompt 3: refine for brand voice and tighten the CTA.
Common Mistake
Trying to cram research, drafting and refining into a single prompt, producing a mediocre result at every step.
Practical Exercise

Break one complex content task into a 3-step prompt chain.

Lesson Checkpoint

Prompt chaining is useful mainly because it

Module Template

**Marketing Prompt Library** — see Templates Library.

Deepen Your Knowledge
  • •Anthropic's prompt engineering documentation
  • •HubSpot Academy — AI prompt lessons
Module Checkpoint Assessment

Test Your Mastery of Module 42

5 questions · 70% passing threshold required

PreviousMod 41: AI Fundamentals
Next ModuleMod 43: AI Content Production