15 Secret Claude Codes
You open Claude, type “build me a landing page,” and fifteen minutes later you have something polished that solves the wrong problem. Or you ask it to think for you instead of with you. Or you replay the same big decision in your head for days because the scary stuff never surfaced early.
The fix is not faster prompting. It is fifteen short phrases you paste before you let the model run — grouped by when you need them. They are not magic words. They are borrowed from strategy, product, psychology, debate, and decision-making work. Each one changes what Claude optimizes for. I am Mika, building with two people and AI on purpose, so when I say these save hours, I mean hours I would have spent unbuilding the wrong thing or second-guessing a call I never pressure-tested.
Run before every project
Use these five when you are about to commit to a direction. They slow Claude down on purpose so you spend less time building the wrong thing.
FIRST PRINCIPLES
Aristotle’s idea; people like Musk and Thiel popularized it in business. Use it when you are entering a space with an existing playbook. You tell Claude to strip borrowed assumptions and reason up from what must be true, not from what everyone else already built.
Before we design anything, reason from first principles: what are we assuming that might be wrong?JOBS TO BE DONE
Clayton Christensen’s framework. Use it before you get attached to your solution. It shifts the question from “what are we building?” to “what would someone hire this to do?” That reframe alone has killed bad products before they ever shipped.
Use jobs to be done: list what people are hiring this for, rank the jobs, and stress-test whether my idea is the best way to get those jobs done.SWOT
Came out of business schools in the 1960s. Use it before you commit. It puts internal strengths and weaknesses on the same page as external opportunities and threats, so you are not just optimistic. You are honest about whether now is even the right time.
Run a SWOT on this decision, then say clearly what it implies for go vs wait.STEELMAN
From rationalist and debate communities. Use it before you pitch or invest. Instead of attacking your own idea, you build the strongest version of the opposing view. You actually understand what you are up against before you are in too deep.
Steelman the strongest argument against this plan. Then tell me what would change my mind.10-10-10
Suzy Welch’s framework. Use it for any go / no-go decision. You ask how you will feel in ten minutes, ten months, and ten years. It separates decisions driven by anxiety from ones driven by real conviction.
Walk me through 10-10-10 for this decision: how will I feel in 10 minutes, 10 months, and 10 years?For smarter thinking
Most people use AI to go faster. That is not the same as getting smarter. These five get Claude to show its reasoning so you actually understand what is happening and come out sharper than when you started.
STEP BY STEP
This comes from chain-of-thought style work in language models (Wei and colleagues, 2022): asking for explicit intermediate steps improves reasoning on messy tasks. Use it when everything feels knotted or the same call keeps getting kicked down the road. It forces Claude to slow down and lay out the chain. The real snag almost always shows up earlier than the line you are staring at.
Step by step: [paste your situation or question]THINK OUT LOUD
From cognitive psychology on verbal reporting (Ericsson and Simon): externalizing reasoning makes gaps visible. Use it when you need to check that a plan, argument, or explanation actually holds together. You get the reasoning on the page, not just a polished answer, so you can see where the thinking jumped a rail.
Think out loud before you conclude: [what you are trying to decide or defend]FIVE WHYS
Borrowed from Toyota lean practice: ask "why" repeatedly (often five times) to get past the surface failure to what is actually driving it. Use it when something is broken, flaky, or slower than it should be. By the end you are usually looking upstream at a constraint or handoff, not the symptom you started with.
Five whys on this: [describe the bug, delay, or failure]RUBBER DUCK
From software culture: explain the problem simply, as if to a rubber duck. Often articulating it fixes your own understanding; with Claude, you want it to reflect and poke holes until the story is coherent. Use it when you genuinely do not know what is wrong. Clear beats clever.
Rubber duck this with me. What I'm trying to do: [...] What I expected: [...] What happened: [...]INVERSION
Associated with Charlie Munger and older Stoic-style practice: flip the question and ask what would guarantee failure. Use it when you have been grinding forward and need distance. Instead of only asking how to make it work, ask what would make it fail or what you are not allowed to ignore. That tends to surface risks you have been routing around.
Inversion: what would guarantee this fails? [describe the plan, launch, or bet]For big decisions
You are not indecisive. You are tired of guessing while something important is on the line. These five force a fast 360 view so you can commit without replaying the same debate in your head for days.
Key insight: If the only voice in the room wants the decision to feel good, you are not deciding. You are decorating.
PRE-MORTEM
Gary Klein’s pre-mortem (1998) is simple and brutal. Pretend it is 12 months from now and the decision failed. Then work backward and list what went wrong, in story form, not bullet shame. It pulls honest risks out while you are still allowed to say them out loud.
PRE-MORTEM — [paste the decision + context]. It’s 12 months from now and this failed. Work backward from that failure: what went wrong, what did we ignore early, and what would have been the first warning signs?RED TEAM
From military and intelligence work: a group assigned to attack the plan, not polish it. In Claude you are not asking for “feedback.” You are asking what a motivated critic would exploit, where the story falls apart, and what evidence would embarrass you if it went public. Structured opposition, not vibes.
You are my red team. Assume you want this decision to fail. List the specific weaknesses, false assumptions, and angles a motivated critic would use against us. No pep talk.DEVIL’S ADVOCATE
An old Vatican role: someone formally assigned to argue the other side. Use it when you notice you are leaning hard one way and every “question” still lands on the same answer. Ask Claude to defend the opposite choice with the best good faith arguments it can build, including who benefits if you are wrong.
Play devil’s advocate for the option I’m not choosing. Give the strongest case for it, including what I’d regret if my preferred path is wrong.SECOND-ORDER THINKING
Howard Marks talks about second order effects: most people stop at the obvious “if we do X, then Y.” Second order asks what Y triggers next (customers, team morale, competitors, cash timing, reputation). Use this before decisions with downstream effects. The surprises usually live one hop past the first spreadsheet row.
SECOND-ORDER THINKING — [paste the decision + context]. Map first-order outcomes, then second-order effects for each (who reacts, what shifts, what new risks appear). Flag the non-obvious ones.TL;DR
After you have run the other passes, compress the decision into one sentence a busy peer could agree to. If you cannot stand behind that sentence, you do not have clarity yet. That is not permission to spin forever. It is a signal to run one more targeted round (usually second order or red team), then decide.
TL;DR this decision in one sentence I could send to a skeptical cofounder. Then list the top 3 risks I’m accepting if I go ahead anyway.How to use these codes
In Claude or any other LLM, use the same pattern every time: name the code, then paste the thought, topic, or decision you want it applied to. A short label and a few lines of context are enough. Example shape: PRE-MORTEM: [paste your situation].
These are not secret unlocks for Claude. Any capable model already knows these phrases from training on books, blogs, news, and everyday language. You are not hacking the system. You are naming the lens so the model reaches for that pattern instead of sliding into generic advice.
Save this for the next time you catch yourself asking Claude to go fast when you need the right frame, clearer thinking, or a honest 360 before you commit.
Here are some related guides to check out: