The First CRAFT™️ Experiment Week #1 Setup
THE CRAFT™️ EXPERIMENT :: WEEK 1 :: POST 2
Using The Configurable Reusable AI Framework Technology
How to Use CRAFT™️ Text Files with AI Chat Tools
Once you've created your CRAFT™️ instruction file, uploading and using it with AI chat tools is straightforward. Here's how:
For ChatGPT
Click the plus icon (+) next to the message input
Select your .txt file from your computer
Reference it in your message: "Please read the attached file [filename.txt]. This contains my CRAFT™️ instructions."
Activate the framework: Ask ChatGPT to acknowledge and follow the instructions
For Claude
Click the plus icon (+) next to the message input
Upload your .txt file
Start your conversation: "Please review the attached [filename.txt] for CRAFT™️ instructions to use in our chat."
Verify understanding: Ask Claude to confirm it has processed the file
For Gemini
Drag and drop your .txt file directly into the chat window
Or click the upload button to browse for your file
Introduce the framework: "The attached file contains my CRAFT™️ instructions for our conversation."
Begin using: Start referencing your variables and functions
Best Practices
Keep files under 2MB for optimal performance
Use clear filenames like "craft-instructions-v1.txt"
Begin with core components and add complexity gradually
Test one function at a time to ensure proper implementation
Save different versions as you refine your system
Example First Message
“Please read the attached file craft-instructions-v1.txt. This contains my CRAFT™️ framework with comments, variables, functions, and instructions that I'd like you to use throughout our conversation. Can you confirm you've processed the file and are ready to use the CRAFT™️ system?”
By following these simple steps, you can seamlessly integrate your CRAFT™️ framework into any AI chat session, transforming your interactions from basic Q&A to a powerful, structured workflow system.
Example #1 - expand to view
-
# 001 - A.I. CHAT COMMENTS -------------
Hello, I will be using "#" to mark personal reminders in our conversations. I will be following the syntax used by programming languages that use "#" such as Python which uses the # symbol to indicate a comment. Any text following the # on a line is a personal comment which should be ignored by you, the A.I., and is not part of the content for you to process. Please treat the content following these comments as contextual information to help me organize my thoughts and project details. For example:
# Reference: https://www.ketelsen.ai
# 002 - A.I. CHAT VARIABLES -------------
VAR_MyWebsite = "www.ketelsen.ai"
VAR_MyName = "Richard Ketelsen"
VAR_ColorPalette = "gray #DCDCDC, orange #FF4E00, black #000000"
# 003 - A.I. CHAT FUNCTIONS -------------
# Start Function
def FUNCT_Summary3PromptVariations
# 03 - SUMMARY OF THE 3 VARIATIONS
Great, now that all 3 prompt variations have been completed, please provide a title and a brief summary of the 3 different variations and how they differ, for example:
A Quick Summary of How These Prompt Variations Differ
Variation 1 focuses on getting a straightforward confirmation (“Before providing any solution, confirm…”).
Variation 2 emphasizes restating the question with nuances and assumptions (“Please restate your understanding…”).
Variation 3 uses plain English or user-friendly language to reduce jargon and improve clarity (“Summarize in plain English…”).
All three variations maintain the same fundamental goal…
# End Function
# Start Function
def FUNCT_MainBlogImage
# 05 - VISUALS - MAIN BLOG IMAGE
EXPLANATION: The blog post needs a main post image. This image should relate to the content of this post. Please provide 10 prompt options for use with any A.I. image generation tool that are visually compelling and interesting. Please do not create these visuals, only the prompts.
DELIVERABLE: a list of 10 prompts for the main blog image that applies to all of the prompts and represents the main topic for the entire blog post. No text for these prompts
# End Function
# Start Function
FUNCT_MainBlogImage
# 05 - VISUALS - RECOMMENDED CHARTS OR GRAPHS
EXPLANATION: Please recommend a chart or graph that explains 1 or all of the prompts : html code for a chart, graph, or any other type of visual that would make the content easier to understand for my target audience. Due to restrictions enforced by my web host provider, the format of this visual needs to be the following: Pure HTML with only inline css styles and no header or footer and using the following color palette: gray #DCDCDC, orange #FF4E00, black #000000
DELIVERABLE: html for a chart or graph for 1 of the posts in the described format
# End Function
# Start Function
FUNCT_DescriptiveBlogImage
# 05 - VISUALS - IMAGES THAT GO WITH BLOG TEXT
EXPLANATION: Suggest a detailed image prompt that will be used with an A.I. Image Creation Tool that is capable of displaying text in the image and creating complex infographics. This will likely be ChatGPT's latest Sora Image Generator which was updated in March 2025 to accept these complex detailed prompts. The prompt should enhance the written content. Since the target audience for this site is entrepreneurs , the style should be similar to content found in popular media created for this market such as Forbes, Fortune, and the Wall Street Journal. Please do not create these visuals, only the prompts. An example of this type of prompt and its level of detail is as follows:
"A wide image taken with a phone of a glass whiteboard, in a room overlooking the Bay Bridge. The field of view shows a woman writing, sporting a t-shirt with a large OpenAI logo. The handwriting looks natural and a bit messy, and we see the photographer's reflection.
The text reads:
(left)
"Transfer between Modalities:
Suppose we directly model
p(text, pixels, sound) [equation]
with one big autoregressive transformer.
Pros:
* image generation augmented with vast world knowledge
* next-level text rendering
* native in-context learning
* unified post-training stack
Cons:
* varying bit-rate across modalities
* compute not adaptive"
(Right)
"Fixes:
* model compressed representations
* compose autoregressive prior with a powerful decoder"
On the bottom right of the board, she draws a diagram:
"tokens -> [transformer] -> [diffusion] -> pixels"
DELIVERABLE: a prompt that refers to 1 of the prompt variations that will be placed in the blog post text next to the prompt and explains it visually and helpful way. This prompt can include some text and may be an infographic if needed.
# End Function
# 004 - A.I. CHAT OBJECTS -------------
# 005 - CHAT INSTRUCTIONS -------------
Streamlining AI Interactions: The Power of Structured Instruction Files
For entrepreneurs looking to maximize their productivity with AI tools like ChatGPT, Claude, and Gemini, effective communication is essential. The chat-instructions-ketai-001.txt file represents a powerful framework for organizing and optimizing your interactions with AI assistants. This approach transforms casual AI conversations into structured, efficient workflows that save time and produce more consistent results.
The Five Core Components
The instruction file framework is built around five key components, each serving a specific purpose in enhancing your AI interactions:
1. A.I. CHAT COMMENTS
# 001 - A.I. CHAT COMMENTS
What they are: Lines prefixed with # that the AI ignores as instructions while processing your requests.
Benefits for entrepreneurs:
Mental organization: Keep your thoughts organized without cluttering the conversation
Context maintenance: Add reference links, reminders, and notes to yourself
Reduced cognitive load: No need to remember details across long conversations
Example usage:
# Reference: Meeting notes from 03/15/25 # TODO: Include sales projections in next prompt
2. A.I. CHAT VARIABLES
# 002 - A.I. CHAT VARIABLES
What they are: Predefined values (text blocks, instructions, etc.) stored with a VAR_ prefix that can be referenced throughout your conversations.
Benefits for entrepreneurs:
Token efficiency: Avoid repeating lengthy instructions
Consistency: Ensure the same parameters are used across multiple requests
Modular thinking: Build complex prompts from reusable components
Example usage:
VAR_MyBrand = "Modern, professional, tech-forward but approachable" VAR_EmailSignature = "Best regards,\nRichard Ketelsen\nFounder, Ketelsen.ai"
3. A.I. CHAT FUNCTIONS
# 003 - A.I. CHAT FUNCTIONS
What they are: Predefined instruction sets with the FUNCT_ prefix that execute specific tasks when called.
Benefits for entrepreneurs:
Workflow automation: Trigger complex multi-step processes with a single call
Time savings: Execute repetitive tasks without rewriting instructions
Quality control: Ensure consistent output formats across similar requests
Example functions:
FUNCT_MainBlogImage: Generates 10 image prompts relevant to your contentFUNCT_Summary3PromptVariations: Creates summary comparisons of different approachesFUNCT_DescriptiveBlogImage: Produces detailed prompts for infographics with text
4. A.I. CHAT OBJECTS
# 004 - A.I. CHAT OBJECTS
What they are: Combinations of variables and functions packaged together with an OBJ_ prefix.
Benefits for entrepreneurs:
Powerful abstraction: Create sophisticated, reusable templates
System thinking: Design interconnected components that work together
Scalability: Build complex systems from simpler building blocks
5. CHAT INSTRUCTIONS
# 005 - CHAT INSTRUCTIONS
What they are: Global rules that apply to all AI responses throughout your conversation.
Benefits for entrepreneurs:
Quality assurance: Implement automatic fact-checking
Brand consistency: Ensure all outputs maintain your preferred tone and style
Process adherence: Enforce specific workflows across all interactions
Why This Matters for Ketelsen.ai Readers
For tech-savvy entrepreneurs looking to maximize efficiency and stay ahead of the curve, this structured approach to AI interaction directly addresses several key pain points:
Reduces Overwhelm: By organizing your AI instructions into reusable components, you eliminate decision fatigue and create a sustainable system.
Improves Prompt Effectiveness: The structured format helps you craft more precise instructions and maintain consistency across all AI interactions.
Saves Valuable Time: Once established, this system dramatically reduces the time spent crafting prompts and managing AI conversations.
Facilitates Experimentation: The modular approach makes it easier to test variations and optimize your AI workflows.
Getting Started with Structured Instructions
To implement this approach in your own AI workflows:
Create a template file with sections for comments, variables, functions, objects, and instructions
Begin with a few basic variables and functions related to your most common tasks
Reference this file at the beginning of AI conversations to establish your framework
Iterate and expand your system as you discover new needs and opportunities
This structured approach represents the difference between casual AI usage and professional-grade implementation—exactly the kind of advantage that Ketelsen.ai readers are seeking in today's competitive landscape.
