Crack the Prompt
The One Prompt Pattern That Always Works
Context → Task → Constraints → Output. Learn it once, use it forever.
If you learn nothing else about prompting, learn this structure: Context → Task → Constraints → Output.
The pattern
Every good prompt has four parts:
- Context — Who are you helping? What’s the situation?
- Task — What do you want the AI to do?
- Constraints — What should it avoid or include?
- Output — What format should the response take?
Example: Before and after
Bad prompt:
“Explain AI”
Good prompt using the pattern:
Same model. Massively better output.
Why this works
AI doesn’t read your mind. It doesn’t know:
- Who you are
- Why you’re asking
- What you’re trying to do
- What “good” looks like to you
If you don’t say it, the AI will guess. And it will guess confidently.
The pattern forces you to answer these questions upfront.
More examples
Marketing email
Context: You’re writing for a SaaS company that sells project management software.
Task: Write an email announcing a new feature (AI task suggestions).
Constraints: Friendly but professional tone. No hype words like “revolutionary.”
Output: Subject line + 150-word email body.
Code review
Context: You’re a senior developer reviewing a junior’s Python code.
Task: Identify potential bugs and suggest improvements.
Constraints: Focus on readability and performance. Be constructive, not harsh.
Output: Numbered list of issues with suggested fixes.
The biggest myth about prompting
You don’t need to be clever. You need to be specific.
Good prompts feel boring. They read like instructions, not magic spells.
That’s a feature, not a flaw.
What’s next?
Want ready-to-use templates? Grab the prompt templates you can steal and customize.