The Founder's AI Prompts Cheat Sheet
Not another '100 ChatGPT prompts' listicle. Six meta-patterns that change how any AI behaves, the per-model quirks that actually matter (Claude wants XML, ChatGPT wants role + context, Gemini wants structure, Perplexity wants citations), and 12 daily founder workflows you'll use this week.
Most founders use AI like a smart Google. The founders we work with who get 10x results from the same models use a handful of meta-patterns I have not seen written down anywhere in one place. This cheat sheet is the working version we hand to every WeActive8 client who asks how we get usable output on the first try. Same patterns we use internally on marketing audits, sales emails, ICP scoring, content briefs, and competitive teardowns. Print it, pin it, run the prompts as written. The point is not to learn AI in the abstract. The point is to ship work today that you could not ship yesterday.
- The 6 meta-patterns: critique-then-revise, pre-mortem, counterparty role, calibrated confidence, cap-the-tokens, anti-sycophancy. One page each, with the copy-paste prompt template and why it works.
- Per-model quirks for Claude / ChatGPT / Gemini / Perplexity — what each one wins at, what each one collapses on, the prompt format each one prefers.
- 12 daily founder workflows: write a cold email that converts, score a lead in 30 seconds, draft a Loom replacement, audit a competitor's pricing page, write a job posting that filters, build a CRO test hypothesis, plus 6 more.
- The 5 prompt patterns NOT to copy — common bad habits that look productive but produce slop.
- A one-page quick-ref cheat sheet at the back, designed to print on a single A4 and pin near your desk.
