Most founders make high-stakes decisions alone, with incomplete information. These seven commands give you a rigorous thinking partner for every strategic choice you face.
A board of directors exists to do one thing: pressure-test your thinking before it costs you. They ask the questions you are too close to the problem to ask yourself. They argue the other side. They surface the risks you glossed over and find the tradeoffs you hoped no one would notice.
Most founders do not have a board. They have a gut feeling, a trusted friend who agrees with them, and a calendar full of decisions that need to be made today. That combination is how smart people make expensive mistakes.
These seven decision codes are prompt structures you can drop into Claude or ChatGPT before a pricing change, a launch, a contract, or a hire. Use them any time the stakes are high enough that being wrong would hurt.
First-principles thinking means removing every assumption until you are left with only what is actually true. Most business decisions get made on top of layers of "that's how we've always done it." This command removes all of that and forces the AI to rebuild the logic from scratch.
Best used before a major pricing decision, before rethinking your offer structure, or any time you are about to do something "because everyone else does it."
/firstprinciples — Strip this problem to its physical roots. Remove every assumption. What is actually true here, and what am I just believing because it feels familiar? Rebuild the logic from scratch.
This command asks the AI to act as a board of directors reviewing your plan. Not a supportive mentor. A group of experienced, financially-motivated people who will not let sloppy thinking through. Give it your plan, your pitch, or your next big idea.
Best used before you pitch something, launch something, or commit significant budget to a direction.
/boardroom — You are my board of directors. Review this plan with rigor. What are the gaps, the assumptions that need more evidence, and the places where my logic breaks down? Do not soften the feedback.
Every email, proposal, or conversation has a subtext. What the other party actually wants, what they are avoiding, what they are testing you on. This command asks the AI to read between the lines and tell you what is not being stated directly.
Best used when you have received a response from a client, investor, or partner that felt off, or when you are trying to decode the real dynamics behind a negotiation.
/coldread — Read between the lines. What is actually being communicated here that is not stated directly? What does this person or situation actually want, and what am I not seeing?
The steelman is the opposite of a straw man. Instead of knocking down a weak argument, you build the strongest possible case against your own position. If you can beat the steelman, your decision is solid. If you cannot, you have found the real objection.
Best used when you are emotionally committed to a decision and need an honest reality check, or before you take a position publicly that you will need to defend.
/steelman — Argue the opposite case better than any critic could. Give me the strongest possible argument against my plan, my decision, or my position. Do not hedge. Make the case as compelling as possible.
Every plan has blind spots. The risks you do not see are the ones that get you. This command surfaces the top five risks specific to your plan, not generic business risks. Works best when you give the AI real context: your timeline, your resources, and your constraints.
Best used mid-planning when everything feels good, or just before you make a commitment you cannot easily reverse.
/risks — What are the top five risks I am not seeing yet? Be specific. Do not list generic business risks. Find the ones that are particular to this plan, this market, and this situation.
Every decision involves giving something up. Founders tend to evaluate the option they are already leaning toward and underweight what they are leaving behind. This command forces a clean side-by-side comparison before you commit, including the hidden costs and second-order effects.
Best used when you are choosing between two or more real options, like whether to hire vs. automate, raise prices vs. expand volume, or niche down vs. stay broad.
/tradeoffs — List the tradeoffs of each option side by side. What do I gain and what do I give up with each path? Include the hidden costs and the second-order effects, not just the obvious ones.
Once you have worked through a decision and have a lot of context on the table, the question is not "what are my options" but "given everything I know, what should I do right now?" This command asks the AI to synthesize everything and give you one clear recommendation.
Best used when you are stuck in analysis paralysis and need a push toward action.
/next — Given everything in this conversation, what is the single highest-leverage move I can make right now? Not a list of options. One clear recommendation with the reasoning behind it.
You now have seven ways to pressure-test any decision before it costs you. Build them into your actual process: /firstprinciples before you build anything new, /boardroom before you pitch, /risks before you commit budget, and /next when you are ready to move. The founders who make consistently good decisions are not smarter. They are more rigorous. These codes give you the rigor.
If you want a system that applies this kind of strategic thinking across your entire business, not just individual decisions, that is exactly what the Decision Engine™ at Her AI Systems™ is built for.
Free Download
Your future self will thank you for saving this.
Follow along for the next guide
New free guide every Tuesday. Find them first on Instagram, TikTok, or by subscribing to the email list.
Want more?
One new free guide. Every Tuesday. No spam, no fluff, just the systems and prompts I'm building for clients, simplified for you.
Subscribe