A fill-in-the-blanks setup for turning any AI into a sharp thinking partner instead of a flatterer.
Most people get bland, agreeable answers from AI because they never tell it who they are or how to behave. This pack fixes that. Fill in the blanks below once, load it into your AI's settings, and every conversation starts from your context with the gloves off.
It works with Claude and ChatGPT. Claude is the recommended tool because it holds the full instructions without limits. ChatGPT works too, on a condensed version, covered below.
Want the fuller picture? Watch the AI Foundations workshop replay for a deeper understanding of how to use this resource.
Using Claude Cowork for file-based work too? Set Up Claude Cowork Properly gives it a memory and a folder system on top of this.
Your instructions have four jobs:
Fill in the fields below. The framework and hard rules are already written for you. As you type, your finished instructions build in the code block below. Then copy them into your AI's settings. You only do this once.
Anything still in [brackets] hasn't been filled in yet. Fill it, or delete it before you paste.
ChatGPT caps each field at about 1,500 characters. The counters below turn red if you go over.
Go to Settings, then your personal preferences (the box asking what Claude should know about you and how it should respond). Copy the Claude block above and paste it in as one block. Done once, it applies to every chat.
Open Settings, then Personalization, then Customize ChatGPT. Paste the "About you" block into the field asking about your background, and the "How to respond" block into the field asking how ChatGPT should respond. Match by what each field asks for, the labels change over time.
We run the whole LinkedIn system, profile, content, lead magnets, outreach, in your real voice. We only take on people we can get results for.
See if we're a fit