AI Prompt Frameworks Cheatsheet

OpenAI Framework

G-CEO Framework

Yeah, But How?

Show Your Work

HTG

Bezos in the Corner

Magical Prompts

OpenAI Framework

Core best practices from OpenAI for crafting effective prompts.

Key Principles:

1

Give Clear Instructions: Be specific about what you want.
2

Provide Reference Text: Offer context, documents, or data for the AI to draw from.
3

Split Complex Tasks: Break down large tasks into smaller, manageable sub-tasks.
4

Give AI "Time to Think": Use phrases like "think step-by-step" or "take a deep breath" for more complex reasoning.
5

Use Examples (Few-Shot Prompting): Show the AI what a good output looks like.
6

Specify Output Format: Clearly state if you want a list, table, email, code, etc.

Pro Tip!

The OpenAI Framework works best when you combine multiple principles. Start with clear instructions and an output format, then add examples or reference text as needed.

BEST FOR:

General good practice, structuring most types of prompts.

G-CEO Framework

A structured approach to ensure your prompt is comprehensive and guides the AI effectively.

G

Goal

Clearly state what you are trying to achieve.

C

Context

Provide all relevant background information, who the audience is, the situation, etc.

E

Examples

If possible, give an example of the desired output or style.

O

Output

Specify the desired format, length, tone, and structure of the AI's response.

Try this format:

Goal: I need to create a presentation about...
Context: This will be presented to a team of...
Example: I like presentations that have...
Output: Please provide a detailed outline with...

Pro Tip!

The more Context you provide, the better! Don't hold back on details that might help the AI understand your needs.

BEST FOR:

Crafting detailed prompts for specific, important tasks where clarity and thoroughness are key.

Yeah, But How?

A very simple but powerful prompt used when you're stuck or the AI gives a high-level answer and you need actionable steps.

The Problem

AI often gives you the "what" but not the "how". You're left knowing what to do, but not how to actually do it.

How to use:

AI SAYS:

"You should create a content calendar to organize your social media strategy."

YOU SAY:

"Yeah, but how do I actually create a content calendar?"

AI SAYS:

"Implement a customer feedback loop to improve your service."

YOU SAY:

"How would I actually set that up step by step?"

Variations to try:

"How would I actually do that?"
"Can you break that down into specific steps?"
"What's the first action I should take?"
"If I've never done this before, how would I start?"

Pro Tip!

This simple prompt can turn abstract advice into concrete action plans. Don't hesitate to ask "Yeah, but how?" multiple times to dig deeper!

BEST FOR:

Getting unstuck, moving from theory to practice, when AI responses are too abstract.

Show Your Work (Chain of Thought)

A technique where you ask the AI to explain its reasoning process step-by-step.

How to use:

Include phrases like:

"Show your work"

"Explain your reasoning step-by-step"

"Break this down logically"

"Think this through"

Example:

INSTEAD OF:

"What's the best investment option for me?"

TRY:

"What's the best investment option for me? Show your work and explain the pros and cons of each option you consider."

Key benefits:

Improves reliability: Forces the AI to be more thorough
Surfaces assumptions: You can see the AI's thought process
Enables correction: You can identify where the AI might be going wrong
Builds confidence: You understand why the AI reached its conclusion

Pro Tip!

This technique works exceptionally well for complex problems involving math, logic, or multiple variables. The AI might still make mistakes, but you'll be able to spot them more easily!

BEST FOR:

Complex problem-solving, calculations, logical reasoning tasks, when you need to understand the "why" behind an answer.

HTG (Hi. Thanks. Great.)

A conversational framework to improve your interaction with the AI and its understanding of your needs.

HI

Start conversationally (e.g., "Hi," "Hello").

Reminds your brain it's a dialogue, not a search query.

THANKS

Acknowledge helpful responses.

Reinforces linguistic matching.

GREAT

Provide specific feedback on the AI's output.

Helps AI understand your preferences.

Example conversation:

You: Hi, I need some ideas for my team meeting tomorrow.

AI: [Provides several meeting ideas]

You: Thanks for those ideas! I like the second one about breakout discussions. Could you expand on how to structure that?

AI: [Provides details on breakout structure]

You: Great, that approach makes sense. I think I'll need some specific questions to guide the breakouts. Could you suggest 3-4?

Pro Tip!

The "Great" feedback is most powerful when it's specific. Tell the AI exactly what you like or don't like about its response, and it will learn your preferences quickly.

BEST FOR:

All interactions to maintain a collaborative dynamic and help the AI learn your preferences over a conversation.

Bezos in the Corner

A thought experiment to remind you of the AI's vast potential as a knowledgeable partner.

The Concept:

Imagine a world-class expert (like Jeff Bezos, or any genius relevant to your field) is always available in the room ("in the corner"), ready to assist. You wouldn't ignore them.

This mindset shift helps you:

Remember to involve AI in your everyday work
Think of AI as a collaborative partner, not just a tool
Leverage its knowledge even when you don't have specific questions

How to use:

Keep your AI tool open

Especially during team meetings or when working on complex tasks

Bring it into conversations

"Let me ask our AI what it thinks about this approach..."

Ask for its "perspective"

Even if you think you know the answer, AI might offer unexpected insights

Pro Tip!

Try asking: "What's a perspective on this problem that we might be missing?" or "How would [famous expert in your field] approach this challenge?"

BEST FOR:

Brainstorming sessions, team meetings, strategic thinking, anytime you might benefit from an "expert" perspective.

Magical Prompts

A collection of creative and unconventional prompting techniques to elicit unique, insightful, or highly tailored responses from the AI.

Split

Have the AI write a pitch, then act as a stubborn customer critiquing it, then have the AI refine the pitch based on that critique. Iterate.

            Try: "First, write me a sales pitch for X. Then, act as a skeptical customer and critique this pitch. Finally, improve the pitch based on these critiques."
          

Bionic

Ask the AI to improve its own previous response.

Try: "Now give me a better answer" or "Make this more persuasive."

Conjure

Ask the AI to adopt the persona of a specific expert or famous figure.

            Try: "Explain this like Einstein would" or "Give me motivational advice like Tony Robbins."
          

Time Travel

Have the AI act as if it's looking back from the future to critique a plan or strategy, identifying potential pitfalls.

            Try: "You are a time-traveling robot. Examine this proposal and tell me what went wrong in 3 years."
          

Mimic

Provide a sample of text (e.g., your own writing, a specific document style) and ask the AI to match that style after you provide your initial rough draft.

            Try: "Here's a sample of my writing style: [paste example]. Now rewrite this draft to match my style: [paste draft]."
          

Pro Tip!

Magical prompts work best when you're willing to iterate and experiment. Don't be afraid to try unusual combinations or to adapt these techniques to your specific needs!

BEST FOR:

Creative tasks, deep strategic thinking, overcoming biases, getting highly specific outputs, and adding a fun, powerful dimension to AI interaction.