Knowledge Publication · 2026
An applied guide to writing and refining prompts, turning AI into a daily production tool for work and decision-making.
About This Publication
A free knowledge publication that teaches one thing clearly: how to turn what is in your head into a precise instruction that is executed as you intended. It starts from a simple rule — do not let AI assume what you want, define it — and builds a full path from user to leader.
The content spans five parts. Part one lays the foundation: understanding AI to the degree you actually need, and dissecting a professional prompt. Part two covers writing and improvement: writing a prompt from idea to text, refining it through diagnosis and iteration, prompt engineering techniques, and context engineering.
Part three is devoted to platforms and configuration: custom instructions, and working across ChatGPT, Claude and Gemini. Part four is application: target sectors with worked examples, and a case study of a prompt system for a full campaign. Part five covers tools and library: an atlas of AI tools, a ready-to-use template library, sources and readings, and a closing chapter on what remains after the tools change.
Key Themes
- 01Introduction: one task, three levels — from user to leader
- 02Understand AI to the degree you need
- 03Dissecting a professional prompt
- 04How to write a prompt: from idea to text
- 05Improving the prompt: diagnosis, iteration and prompt engineering techniques
- 06Context engineering and custom instruction setup
- 07The three platforms: ChatGPT · Claude · Gemini
- 08Application: target sectors and a case study of a full-campaign prompt system
Key Insights
- Do not let AI assume what you want. Define it.
- One goal: turning what is in your head into a precise instruction executed as intended.
- Moving from tool user to leader passes through three levels of a single task.
- A professional prompt is a structure that can be dissected, not a lucky sentence.
- Improvement comes from diagnosis and iteration, not random rewording.
- Context engineering and custom instructions matter more than switching tools.
- Each platform behaves differently: ChatGPT, Claude and Gemini are not led the same way.
- What remains after the tools change is how you think and how you frame the request.
About Mohammed
Board Member & Head of Sales and Marketing Sector at Waheej Real Estate, working across marketing and sales leadership, investment and opportunity assessment, product development, pricing, feasibility and market research — pairing executive governance with market insight, Artificial Intelligence and PropTech.
Riyadh, Saudi Arabia