Ten universal commands that work with any AI model and any topic. Save them once, paste them daily: clarify before answering, critique your own work, distill a long doc, rewrite in another voice, run a pre-mortem, end the day with a structured review.
Anyone who paste-talks to AI more than five times a day - founders, marketers, PMs, writers, support leads, students. These prompts work the same whether you use ChatGPT, Claude, or Gemini, and whether the topic is code, copy, strategy, or homework. If you have a chat-history graveyard of prompts you wrote once and never found again, this pack is the cure.
Most AI advice teaches you to write better prompts. This pack flips that: it gives you ten short commands you write once and reuse forever. They are the meta-moves that compound - clarify before answering, critique your own draft, distill a long doc, run a pre-mortem, end the day with a structured review. Saving them in your clipboard means the next ten thousand AI conversations get instantly better with no extra typing.
Tap any prompt to copy it now, or add all 10 to PromptPaste for one-tap access anywhere. Variables like {{language}} become fillable fields inside the app.
Before answering my next message, ask me up to 5 short clarifying questions about audience, length, tone, format, and any constraints I should have specified. Do not attempt the task yet. Wait for my answers, then proceed.
Why it works: Models default to confidently answering whatever they think you asked.
Example: Before I answer, I need to clarify a few things: 1.
Read the following document and produce:
1. A one-line takeaway (the single most important thing).
2. The three highest-leverage facts (claims, numbers, decisions).
3. An explicit list of what you cut and why.
Do not include filler or background. Treat me as someone who has 60 seconds to decide whether to read the original.
Document:
{{document}}
Why it works: Compression is the most common AI task and the one most people do badly.
Example: One-line takeaway: Postgres 17's logical replication can replace most ETL pipelines for sub-100GB warehouses, but only if you control schema changes.
Explain {{topic}} to {{audience}} in under 100 words. Use only words and analogies they would recognize. No jargon unless you immediately define it. End with one concrete example, not a generalization.
Why it works: Most explanations from AI sit in the awkward middle: too technical for a beginner, too shallow for an expert.
Example: For a smart 12-year-old, in under 100 words: A database index is like the index at the back of a textbook.
Read the draft below as if you were {{skeptic_role}}. Find the specific weaknesses they would push back on. Quote the exact lines you object to. Do not be polite - be useful. Do not suggest fixes; just name the problems clearly enough that I can fix them myself. Draft: {{draft}}
Why it works: Models are sycophantic by default - they'll happily polish a draft that has a fundamental problem.
Example: Reading this as a skeptical CFO, here's what I'd push back on: 1.
Rewrite the source text below in the voice demonstrated by the examples. Then, in a separate section called 'Style notes I applied,' list the 5-8 specific moves you used (sentence length, banned words, openings, anything else you noticed in the examples). Examples of the target voice: {{voice_examples}} Source text to rewrite: {{source_text}}
Why it works: Voice transfer is one of the few things AI does well and most people use badly.
Example: Rewritten in the target voice: We shipped Postgres 17 support today.
I have a decision to make: {{decision}}. Argue the case from each of these perspectives, one at a time: {{perspectives}} Then produce a 'conflict map' showing where the perspectives actually disagree, and a synthesized recommendation that takes the disagreement seriously - not one that smooths it over.
Why it works: Decisions made by a single mind have a single set of blind spots.
Example: The decision: charge $59/yr or $99/yr for the annual plan.
Six months from now, this project failed. Work backward and produce three layers:
1. **Causal failure:** the proximate reason it failed.
2. **Upstream cause:** what was true earlier that made the proximate failure inevitable.
3. **Early warning sign:** something I could observe today that would predict this failure.
Then propose one mitigation I could implement this week.
The project:
{{project_description}}
Why it works: Pre-mortem is the single most under-used decision tool in modern work.
Example: Six months from now, the launch failed.
Help me close the day. Ask me four short questions, one at a time, and wait for my answer before moving to the next: 1. What actually shipped today (not what you intended)? 2. What stalled, and what is the specific blocker? 3. What surprised you - a signal you didn't expect? 4. What does tomorrow's you need to know in one sentence? When I've answered all four, format the responses as a dated journal entry I can save.
Why it works: End-of-day reviews are the highest-leverage daily ritual in knowledge work, and the one most people skip because they don't have a structure.
Example: End of day, 2026-04-26: Shipped: seo audit doc, marketing pack research, power prompts pack draft.
Read the following piece and identify the single load-bearing claim - the one that, if false, makes the rest of the argument collapse. State it in one sentence.
Then build the strongest possible argument against that claim. Cite specific counter-evidence, alternative framings, or assumptions the piece is making but not defending.
End with a one-line verdict: does the claim survive the attack, and what would the piece need to add for it to survive better?
The piece:
{{piece}}
Why it works: Most AI outputs are fine prose with one weak claim that nobody notices.
Example: Load-bearing claim: 'Marketers will install a Mac/iOS clipboard app to save AI prompts.' Strongest argument against: Marketers already have ChatGPT 'Saved Prompts,' Claude…
Below are 3-5 examples of writing in the voice I want to capture. Read all of them. Then produce a voice guide with:
- 5-8 specific moves you noticed (sentence length, banned words, openings, tone shifts, formatting habits).
- One short paragraph describing the overall feel.
- A one-line 'do not do' list of moves that would break the voice.
Make the guide short enough that I can paste it as a prefix to any future prompt.
Examples:
{{voice_examples}}
Why it works: Brand voice is the most reused asset in marketing work and the most awkward to capture.
Example: Voice guide extracted from the 4 examples: Sentences are short.
Ten universal commands that work with any AI model and any topic. Save them once, paste them daily: clarify before answering, critique your own work, distill a long doc, rewrite in another voice, run a pre-mortem, end the day with a structured review. Open the pack in PromptPaste and all 10 prompts import as a single folder you can use instantly.
Because the bottleneck has shifted. Models are smart enough that the limiting factor is the question, not the answer. A meta-prompt that forces the model to ask clarifying questions, critique its own draft, or surface its assumptions adds more value in one paste than rewriting your whole prompt does. The same ten commands work across ChatGPT, Claude, and Gemini, and across every topic you'll ever ask about - that's why they belong in your clipboard, not your chat history.
Yes. Prompts copy to your clipboard as plain text, so they work with any AI model — ChatGPT, Claude, Gemini, or any other.
Eight tested prompts for the marketing tasks you do weekly: subject lines, blog outlines, ad copy, repurposing, cold outreach, brand-voice extraction, customer pain-point synthesis. Each ships with rationale, an example output, and the common mistake to avoid.
Eighteen tested prompts for the PM tasks you do every week: PRD drafts, interview synthesis, stakeholder updates, decision memos, sprint retros, Jira tickets, A/B test hypotheses - plus a four-step chain that takes raw customer feedback all the way to a decision memo.
Seventeen tested prompts for the recurring tasks of running a one-person business: cold outreach, customer onboarding, scope replies, invoice chases, weekly content, pricing - plus a four-step chain from a lead to a sent proposal.