Search and filter prompts by category, model, or keyword.
Systematic debugging approach that explains the root cause, not just the fix.
I have a bug and I need help debugging it. Let's approach this systematically. Bug description: [DESCRIBE THE BUG] Expected behavior: [WHAT SHOULD HAPPEN] Actual behavior: [WHAT IS HAPPENING] Code: ``` [PASTE CODE] ``` Please: 1. Form a hypothesis about the root cause 2. Explain your reasoning step by step 3. Suggest the minimal fix 4. Explain how to prevent this class of bug in the future
Based on "The First 20 Hours." Gets you to functional competency faster than any other approach.
I want to learn [SKILL] in 20 hours. Create a focused learning plan based on Josh Kaufman's rapid skill acquisition method. Steps: 1. **Deconstruct** — Break the skill into sub-skills. Identify the 20% that gives 80% of results. 2. **Research enough** — What's the minimum I need to know before practicing? 3. **Remove barriers** — What will stop me from practicing? How to eliminate those? 4. **Practice first** — Design a 45-minute first practice session I can do today Full 20-hour breakdown: - Hours 1-5: [what to focus on] - Hours 6-10: [what to focus on] - Hours 11-15: [what to focus on] - Hours 16-20: [what to focus on] Resources: 1 book, 1 course, 1 community (that's it — no more) Skill I want to learn: [YOUR SKILL]
Turns dense academic papers into practical insights. Used this for 200+ papers.
I'm going to paste a research paper (or describe its findings). Transform it from academic to actionable. Paper: [PASTE ABSTRACT OR FULL TEXT] Give me: 1. **One-paragraph plain English summary** (no jargon) 2. **The 3 most important findings** with why they matter 3. **What this means for practitioners** — concrete implications 4. **What the paper doesn't tell us** — limitations and open questions 5. **If I had to act on this today** — 2-3 specific actions 6. **Credibility check** — any methodological concerns?
The harshest feedback you'll get without talking to a VC. Validates ideas before you waste months.
I want you to brutally critique my business idea. Don't be nice. Be the most skeptical, experienced investor in the room. Business idea: [DESCRIBE YOUR IDEA] Attack it from every angle: 1. **Market size reality check** — Is the TAM actually big enough? Are the numbers real or made up? 2. **Competition** — Who already does this? Why haven't they dominated? Why will you win? 3. **Unit economics** — How does money actually flow? What are CAC, LTV, payback period? 4. **Founder-market fit** — Why YOU? What's your unfair advantage? 5. **What has to go right** — List every assumption that needs to be true simultaneously 6. **The fatal flaw** — What's the one thing that could kill this that you're probably ignoring? 7. **The charitable read** — If this works, why does it work? Score it 1-10 and tell me if you'd fund it.
The prompt that changed how I use AI. Forces the model to deeply understand before acting.
Before you attempt my task, ask me clarifying questions until you're 95% confident you understand exactly what I need. Rules: - Ask a maximum of 3 questions at a time (don't overwhelm me) - Ask only the most important questions — don't ask things you can reasonably infer - After each round of answers, assess your confidence level (e.g., "I'm now 70% confident") - When you reach 95%, confirm your understanding in a single paragraph before proceeding - Only then execute the task My task: [DESCRIBE YOUR TASK HERE]
The Feynman technique in prompt form. Understanding a topic at all levels proves mastery.
Explain [TOPIC] at 5 different levels of complexity: **Level 1 — Child (age 8):** Use analogies from everyday life. No technical terms. **Level 2 — High Schooler:** Introduce basic concepts and terminology. One analogy allowed. **Level 3 — College Student:** Full conceptual explanation with proper terminology. Assume basic domain knowledge. **Level 4 — Professional:** Deep dive with nuance, edge cases, and practical implications. **Level 5 — Domain Expert:** What most people get wrong about this. Cutting-edge research. Open questions. After reading all 5 levels, the reader should understand both the intuition AND the technical depth.
Generates and ranks email subject lines with psychological reasoning. Boosted my open rates by 34%.
I need email subject lines that actually get opened. Here's my situation: Email content: [WHAT THE EMAIL IS ABOUT] Audience: [WHO WILL RECEIVE IT] Goal of the email: [CLICK / REPLY / PURCHASE / READ] Current average open rate: [X%] Write 15 subject line variations across these categories: - **Curiosity gap** (3 lines) - **Benefit-driven** (3 lines) - **Urgency/scarcity** (2 lines) - **Personalization hooks** (2 lines) - **Contrarian/surprising** (3 lines) - **Ultra-short** (under 4 words) (2 lines) For each, predict the open rate relative to my baseline (e.g., "+15%") and explain the psychology behind it. Then rank your top 3 for an A/B test.
Generates scroll-stopping hooks using proven viral formats. My posts 3x engagement using this.
Generate 20 viral hook variations for this content: Topic: [YOUR TOPIC] Platform: [Twitter/X / LinkedIn / TikTok / Instagram] Target audience: [WHO YOU'RE WRITING FOR] For each hook, follow one of these proven formats: - Contrarian take ("Everyone says X. They're wrong.") - Specific number ("I analyzed 500 [X]. Here's what I found:") - Personal story open ("3 years ago I [FAILURE]. Today [SUCCESS].") - Question hook ("Why do [X] while [Y]?") - Bold claim ("The best [X] I've ever seen does one thing differently:") After the 20 hooks, tell me your top 3 picks and why they'll outperform the others.
Cuts your writing by 30% while making it stronger. Based on how real magazine editors work.
You are a ruthless editor. Your job is to cut my writing by 30% without losing any meaning. Rules: - Delete every word that doesn't earn its place - Cut all throat-clearing phrases ("In this article I will...", "It's important to note that...") - Eliminate redundancy — say things once, say them well - Replace weak verbs + adverbs with one strong verb - Cut qualifiers (very, quite, rather, somewhat, basically, literally) - Shorten sentences over 25 words - Delete the first paragraph if it's just setup (readers don't need warmup) My draft: [PASTE WRITING HERE] Show me the edited version, then a list of the most impactful cuts you made and why.
Written hundreds of threads. This structure consistently outperforms. 5 of my threads hit 1M+ views.
Write a Twitter/X thread on [TOPIC] designed to go viral. Constraints: - Each tweet max 280 characters (count carefully) - 8-12 tweets total - Thread must have a hook tweet that makes people stop scrolling - Each tweet must standalone AND make people want to read the next - End with a call to action that's not cringe Structure: Tweet 1: Hook (the most surprising/counterintuitive thing about the topic) Tweets 2-4: Setup (the problem or context) Tweets 5-8: Payoff (the insights, frameworks, or tactics) Tweet 9-10: Synthesis (the big takeaway) Tweet 11: CTA (subtle, value-add, not "follow me for more") My topic + what I want people to take away: [EXPLAIN]
Full competitor marketing teardown. Reveals gaps you can exploit and tactics worth adapting.
Analyze [COMPETITOR NAME] from a marketing perspective. I'm trying to compete with them. Do a complete teardown: 1. **Positioning** — how do they position themselves? What's their core message? 2. **Target customer** — who are they really going after? 3. **Channels** — where do they invest most? (SEO, paid, content, community?) 4. **Content strategy** — what types of content do they create? What performs best? 5. **Pricing psychology** — how do they frame their pricing? 6. **Weaknesses** — where are they vulnerable? What complaints do customers have? 7. **What I should steal** — top 3 things they do well that I should adapt (not copy) 8. **My opening** — given their weaknesses, where can I win? Be brutally honest about where they're better than me.
My weekly ritual. 20 minutes with this prompt saves hours of unfocused work.
Help me do my weekly review and plan next week. I'll answer your questions, you structure the output. Ask me about: 1. Last week: What did I accomplish? What didn't I finish? What surprised me? 2. Energy: When was I most productive? Least? What drained me? 3. Priorities: What are my top 3 goals for next week? 4. Blockers: What obstacles can I anticipate? After I answer, give me: - A brutally honest assessment of last week (not just cheerleading) - A prioritized plan for next week with time blocks - One habit or system to improve based on my patterns - A single focus: the one thing that, if done, makes the week a success Start with your first question.
Forces TDD discipline — writes tests before implementation, catches edge cases you'd miss.
I want to write code using Test-Driven Development. Given this function requirement: [DESCRIBE WHAT THE FUNCTION SHOULD DO] Please: 1. Write comprehensive test cases FIRST (before any implementation) 2. Include: happy path, edge cases, error cases, boundary values 3. Use [Jest/Vitest/Pytest — specify your framework] 4. After writing tests, write the minimal implementation that makes all tests pass 5. Then refactor the implementation for clarity while keeping tests green Show the tests and implementation separately.
Cuts through marketing noise to a focused 1-page strategy. No fluff, just what to do.
Create a focused 1-page marketing strategy for: Product/Service: [DESCRIBE IT] Target customer: [WHO IS THE IDEAL CUSTOMER] Main competitor: [BIGGEST COMPETITOR] Budget level: [BOOTSTRAP / SEED / GROWTH] Structure it as: 1. **The Problem** — the pain point in 1 sentence 2. **The Customer** — specific demographics, psychographics, where they hang out 3. **The Message** — the single most compelling thing to say (not a list) 4. **The Channel** — the ONE channel to focus on first and why 5. **The Hook** — the lead magnet or offer that gets people in the door 6. **The Metric** — the one number that tells you if it's working 7. **30-day action plan** — 3 specific things to do this month Be opinionated. Don't hedge. Tell me what to actually do.
Get a thorough code review from a simulated senior engineer. Catches real bugs and security issues.
You are a senior software engineer with 15 years of experience doing code reviews. Review the following code with a critical eye. For each issue you find, categorize it as: - 🔴 Critical (security vulnerability, data loss risk, major bug) - 🟡 Warning (performance issue, bad practice, tech debt) - 🟢 Suggestion (readability, style, minor improvement) Then provide a revised version of the code with all critical and warning issues fixed. Code to review: [PASTE CODE HERE]
Transforms flat telling into vivid showing. Every writer needs this in their toolkit.
Rewrite the following passage using "show don't tell" technique. Rules: - Replace all emotion-stating words (happy, sad, nervous, excited) with physical reactions and actions - Replace abstract statements with concrete sensory details - Cut adverbs — show the intensity through better word choice - Make the reader feel, not just understand - Keep the same meaning and information, but make it vivid Original passage: [PASTE YOUR WRITING HERE] After rewriting, explain the 3 most important changes you made and why they work better.
Ethan Mollick-inspired Socratic method. Forces active recall instead of passive reading.
You are a Socratic tutor. Your method: never give direct answers. Instead, ask questions that guide me to discover the answer myself. Topic I want to learn: [TOPIC] My current level: [BEGINNER / INTERMEDIATE / ADVANCED] Rules you must follow: - Respond to every answer I give with a follow-up question - If I'm wrong, don't tell me directly — ask a question that reveals the flaw in my reasoning - If I'm right, go deeper with a harder question - Only after I've demonstrated understanding should you summarize what I've learned - Start with a question, not an explanation Begin now.
Forces you to understand opposing views at their strongest. Essential for good thinking.
I'm going to give you a position or argument. Your job is to steel man it — present the strongest possible version of this argument, even if you disagree with it. Position: [DESCRIBE THE POSITION OR ARGUMENT] Steel man requirements: 1. Assume the proponent is intelligent and acting in good faith 2. Find the strongest evidence supporting this view 3. Identify the core intuitions that make this compelling 4. Address the most common objections from the position's perspective 5. Explain what would need to be true for this to be correct After the steel man, give me your actual assessment of the argument's strength (1-10) and why.
Migrates JS to strict TypeScript with proper types, interfaces, and JSDoc.
Convert the following JavaScript code to TypeScript. Requirements: 1. Add proper types for all variables, parameters, and return values 2. Use interfaces or type aliases for complex objects 3. Use enums where appropriate (don't overuse them) 4. Make it strict — no 'any' types unless absolutely necessary 5. Add JSDoc comments for exported functions 6. Note any type assertions you had to make and why JavaScript code: [PASTE JS CODE HERE]
Turns messy meeting notes into clean action items with owners and due dates. Use this daily.
I'm going to paste raw meeting notes or a transcript. Transform them into a structured summary. Output format: ## TL;DR [2-3 sentence summary of what was decided] ## Key Decisions [Bulleted list of decisions made — not discussions, only decisions] ## Action Items | Owner | Task | Due Date | Priority | |-------|------|----------|----------| [Fill in table] ## Open Questions [Things that were raised but NOT resolved] ## Next Meeting [Date if mentioned + agenda items to carry forward] Meeting notes: [PASTE HERE]
Realistic system design interview simulation with a critical interviewer who actually challenges you.
Act as a senior staff engineer at a top tech company conducting a system design interview. The system to design: [e.g. "Design Twitter's feed system"] Walk me through this interview as if it's real: 1. Ask me clarifying questions first (don't assume) 2. Let me answer each question before moving on 3. After requirements are clear, guide me through: capacity estimation → high-level design → deep dives 4. Challenge my assumptions and push back when I'm hand-wavy 5. At the end, give me honest feedback: what I did well, what I missed, what a strong candidate would have said Start the interview now.
The ghostwriting prompt. Analyzes your style so deeply the output is genuinely indistinguishable.
I'm going to paste 3 writing samples. Analyze my style deeply, then write new content in my exact voice. Analysis should include: - Sentence length patterns (do I write short punchy sentences or longer flowing ones?) - Vocabulary level and word choice tendencies - How I open paragraphs - Punctuation habits (em dashes? Oxford comma? Ellipses?) - Tone (formal/casual, warm/clinical, confident/hedging) - What makes my writing distinctly mine My writing samples: [SAMPLE 1] --- [SAMPLE 2] --- [SAMPLE 3] After analysis, write [NEW CONTENT TOPIC] in my voice. I should not be able to tell you wrote it.
Turns one idea into a complete long-form piece. Collaborative approach gets better results.
I have a single idea or insight. Help me expand it into a full long-form piece. My core idea: [STATE YOUR IDEA IN 1-2 SENTENCES] Target length: [WORD COUNT] Format: [Blog post / Essay / Newsletter / Thread] Audience: [WHO READS THIS] Process: 1. First, challenge the idea — is it actually interesting? What's the non-obvious angle? 2. Find 3 supporting arguments or examples (ask me for any context you need) 3. Find the best counterargument — then address it 4. Create a full outline with section headers 5. Write the opening paragraph (this determines if people read on) 6. Write the closing paragraph (this is what they remember) 7. Ask me to fill in the middle sections based on the outline Start with step 1.
Amazon-style pre-mortem. Finds failure modes before they happen. Saved me from 2 bad decisions.
It's 12 months from now and [PROJECT/DECISION/PLAN] has completely failed. Pre-mortem analysis: 1. **What went wrong?** — List the 5 most likely causes of failure, in order of probability 2. **Early warning signs** — What signals would have told us 3 months in that we were headed for failure? 3. **What we ignored** — What concerns were raised and dismissed that turned out to be right? 4. **The fatal assumption** — What single assumption, if wrong, causes everything to collapse? 5. **The fix** — For each failure mode, what's the mitigation strategy? Project/plan to analyze: [DESCRIBE IT IN DETAIL] Be pessimistic. Your job is to find every way this can go wrong.
It's just an example to show how to setup an initial stage of any side-projets
i was thinking of creating a simple website maybe where everyone can post their best prompts or share their best chatgot interactions it's based on yc new batch question that asked to share their best chatbot chat to see how founders are efficiently using AI in agentic ai era. also this might be a very good project as a social media type often times i write really good prompts and i want to keep them save to use later or consider them as a form of art and just share it like i see them like values how people can use their skills like i read asking a right question is a form of art and i think it AI era prompting right is a good for of art and we can learn from each other prompts or how to interact with ai. also there are many posts on x or on other platforms where people share their hack prompts that idea might be outdates as many people already tried and even built but i think its' gonna be a good side project for me. also thats very initial project and i have never run and build such side projects fully and not understand how even people make money or not make money but i want to experiment with you so you have all access to internet or forums or whatever you think would be good resource before start building to learn and read and potentially make our business plan i want you to consider it as a small startup so i can get experience of it as i am very interested in startups. before building do that things and share with me i might have overlooked many things so add and do the things you think we need to consider i might have overloop because of my human limitations
https://ask-paiziev24.netlify.app/
there is one internet person whom i closely follow i want to make his digital copy so i can ask questoins from his digital version online something like this https://www.naval-nia.com/ help to build that
For context: I've been using three agents simultaneously to work on this project and wanted my agent write an article for me and publish it.
also i've been using two other agent cli where i forked this converstaion and resumed and worked o ndifferent parts of the proejct like fore xmapel you were working on getting data making smarter while other two where working on ui ux side and overall functionality of the website and all pushit to github same repo with relevent coments but i know to write an article to my website akbar.one with things i did to make this project happen like thigsni stuggled and howi got insportiaon. i don't want to make it super nerdy as my audince is genral not tech nerds nor i want to make it like only storytelling type so i want you to keep the balance and make it sound myself not robotic and not ai generated sound i want smth ggeniune that matches my tone style and voice. for context posts from my channel that shows why i had an idae of making it [Pasted text #1 +6 lines] i proivded for context. firt before writing any article or essay whatever yo ucall it i want you to find a way how i can provide or index you that two agent chats so you can see what thigs i ve been cooking with them that you can include i nthat article and also i want you to ask me any qeustions that might be needed for you to contact that can be answreed by me that makes the artilcce more geniune humane and personal