Skip to content

Instantly share code, notes, and snippets.

@atran
Last active February 28, 2025 21:15
Show Gist options
  • Save atran/c9ac7c5b4d8a7d60e2256430658c69af to your computer and use it in GitHub Desktop.
Save atran/c9ac7c5b4d8a7d60e2256430658c69af to your computer and use it in GitHub Desktop.
Master Prompt: Applying Core Prompt Techniques to Your Project
I have attached an image of my project, which includes an input, parameters, and an output structure.
Generate a structured list of copy-and-paste prompts that will help me iteratively refine, analyze, and expand my project. These prompts should align with the core techniques listed below. Each should be written as if I am asking an AI model a question to engage with my project more effectively.
Each prompt should:
• Reference my project’s specific input, parameters, and output
• Follow the structure of the associated technique
• Encourage deeper exploration of interactive workflows, system dependencies, or generative design
Provide 14 prompts following this structure:”
Core Prompt Techniques for the AI to Generate Prompts
Basic Techniques
1. Zero-Shot Prompting – Generate a prompt that asks the AI to analyze and explain my project with no prior context.
2. Few-Shot Prompting – Generate a prompt that provides 2-3 example projects and asks the AI to compare or build upon them.
3. Chain-of-Thought Prompting – Generate a prompt that asks the AI to break down the transformation process of my project into sequential steps.
Structural Techniques
4. Instructional Prompting – Generate a prompt that asks the AI to write a step-by-step instructional guide for executing my project.
5. Role-Based Prompting – Generate a prompt that asks the AI to take on the role of a relevant expert (e.g., computational designer, urban planner, sustainability analyst) and provide insights.
6. Delimiting Prompting – Generate a prompt that asks the AI to focus only on a specific aspect of my project, such as usability, performance optimization, or material selection.
Advanced Techniques
7. Debate Prompting – Generate a prompt that challenges the AI to argue against a key assumption in my project and then defend or refine it.
8. Tree-of-Thought Prompting – Generate a prompt that asks the AI to explore multiple design alternatives or workflows for my project.
9. Fan-Out Prompting – Generate a prompt that asks the AI to expand my project into multiple possible applications or integrate it into larger systems.
10. Chaining, Least-to-Most Prompting – Generate a prompt that starts with a basic version of my project concept and progressively introduces more complex elements.
System Techniques
11. Multi-Modal Prompting – Generate a prompt that asks the AI to analyze my project using multiple data types (e.g., images, structured text, numerical data).
12. Reflection Prompting – Generate a prompt that asks the AI to critique its own explanation of my project and suggest improvements.
13. Recursive Criticism – Generate a prompt that asks the AI to iterate on its response by addressing specific weaknesses, missing details, or inconsistencies.
14. Meta-Prompting – Generate a prompt that asks the AI to generate the next best question I should ask to further refine my project.
Deliverables
The AI should return a structured list of 14 copy-and-paste prompts that I can use iteratively to refine my project. Each prompt should encourage deeper interaction with AI-driven workflows.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment