My Prompt Stack for Work: 16 Prompts In My AI Toolkit That Make Work a LOT Easier

These prompts cover everything from presentations to product to analyzing customer data to smart tradeoff thinking—they show how I'm using smart prompts to make AI a true partner at work!

Oops I did it again (wrote an obscenely long treasury article about AI lol). I was also hungry, so the picture is a cute short stack of pancakes.

Seriously, this is 66 pages of prompts, and yes there is a table of contents so you can figure out what you need. It covers your entire work stack, and each prompt is obsessively crafted.

Why these 16 prompts? Because they have coverage across the work stack, and because they’re designed to work together to help you think and partner with AI usefully. I a Substack piece about how it sometimes takes multiple prompts to develop a really great insight with AI—well sometimes you can fast forward that with a great starter prompt. That’s what this article is supposed to do—help fast forward your work stack.

After you read this article, let me know other prompts you’d like me to work on next!

I worked hard on these. I’ve spent more time than I’d like to admit with these models. Not just asking them for help, but pushing them. Stress-testing their edge cases. Breaking them. Rebuilding the way I think to match the way they think. Over time, I developed a kind of fingertip feel for what works—a sense for prompt architecture, for how to bend the model toward clarity without letting it collapse into generic noise.

Each of these 16 prompts is designed not just to elicit better output—but to shape better thinking. They force you to slow down where it matters, to tighten your language, to expose ambiguity in your own mind before it gets mirrored back at you. They treat prompting as what it actually is: a structure of thought. Something that teaches you as much as it teaches the model. Yes, we’re learning too!

I don’t write prompts to delegate thinking. I write prompts to think through. I expect every prompt I run to perform—not just to generate, but to reason, to push back, to clarify, to deliver something shaped, usable, and precise. That means the prompt has to carry weight. It has to hold form in latent space. It has to be built for runnability and iteration. These prompts aren’t speculative. They’ve been run and refined, tested in real workflows, and updated until they snapped into place.

Before we get into them, I want to show you the architecture that makes these prompts work. There’s a pattern to good prompting—structure, context, constraints, role, format, and feedback—and once you internalize that pattern, you stop guessing. You stop hoping the model will figure it out. You start building prompts the same way you build software or strategy: deliberately, with intent.

Let’s start there.